Re: Install Debian 12.5 on QNAP TS-210

[Alexander V. Makartsev]

On 24.04.2024 19:49, David Hörnlund wrote:

Hi debian-user,

I have an old QNAP TS-210 that would continue to be useful for me. If 
it is still possible to use it with the latest Debian Stable.


There is a webpage at https://www.cyrius.com/debian/kirkwood/qnap/ts-219/

That have instruktions on how to install Debian 10 on this device.

Can I install Debian 10 on the device today still or is that version 
not available anymore? Are the .Deb packades still on the mirrors? If 
the files are available will it install or is it blocked somehow?


If I follow the steps on the webpage and do what the author suggests:

My recommendation is for the third option;  Arnaud Mouiche has created 
a script that re-configures the partition layout.


Arnaud Mouiche's method has been used by many users with success.


Will Debian 12 have the necessary packades and cpu architecture 
support to allow Debian 12 to boot and run. But also receive updates 
and new software configured for the QNAP TS-210?



According to the author Debian dropped the support for this device 
when support ended for Debian 10.


The Marvel ARM SoC IC is still supported by kernel [1], so the only 
reason support was ended at Debian 10 is indeed because of limited space 
on the flash device.
I don't see why it shouldn't work, if information on Arnaud Mouiche's 
Github page [2] is correct.

If I had this device I'd definitely gave it a try.
But you have to be careful, since the whole process requires re-flashing 
custom firmware, you might end up with bricked device.

Another option is to simply continue to use QNAP internal OS.


[1] https://www.kernel.org/doc/html/v6.1/arm/marvell.html#kirkwood-family
[2] https://github.com/amouiche/qnap_mtd_resize_for_bullseye

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Root password strength

[Alexander V. Makartsev]

On 22.03.2024 14:57, Jan Krapivin wrote:



чт, 21 мар. 2024 г. в 22:34, Alexander V. Makartsev :

This conclusion seems less than optimal to me.
By condemning yourself to type 12+ character password every time
you 'sudo' would really hurt accessibility and usability of your
home computer and for no good reason.

If we focus solely on your use case: a login security of a PC at
home, without remote access, then password of your sudo user could
be as short and
simple as four numbers, of course unrelated to your date of birth,
phone number, or any other easily guessable sequence of numbers,
like '1234'.


Are you speaking only about sudo or root password also?


Dealing with root password could be tricky and you have three options:
1. You can implement the same 'faillock' scheme for root user as well 
and make root password shorter for convenience.
    Pro: 3 failed login attempts and root user will be locked for a 
time period.
    Con: You or somebody can (un)intentionally lock out root user for a 
time period.
2. You can set good password (12+ symbols) for root user without 
'faillock' scheme.

    Pro: You will be always able to login as root user.
    Con: Typing 12+ symbols password could be a headache.
3. You can unset (delete) root user password and lock the account.
    Pro: Nobody will be able to login as root user directly. Instead 
you will have to rely on sudo user to gain root privileges.
    Con: You will have to keep sudo account safe and set shorter lockup 
time period or make another sudo user as backup.


If you prefer to have root user as failsafe, to fix system when you 
screw something up. I suggest to go for option 2 and keep it simple.


The thing that bothers me are words: "*_any_* computer (and a fortiori 
any server) connected to the Internet*_is regularly targeted by 
automated connection attempts"

_*
I am not tech-savvy. Can you say with 100% (90%?) confidence that 
there is no such thing? That home PC without SSH and whatever 
complicated is safe (rather safe) from "*_automated connection attempts"?

_*
This thread reminded of that topic - 
https://forums.debian.net/viewtopic.php?t=154002

*_
_*
That statement is not entirely true, because it depends on a method how 
a PC is connected to the Internet. There are three options:
1. Your PC is connected to Local Area Network (LAN) and there is a 
router/firewall device between your PC and the Internet cord.
    In this case any unsolicited Internet traffic (automated 
connections, port scans, etc) will be stopped by router/firewall device.
    This is because of how IPv4 network address translation (NAT) 
works, to allow multiple LAN hosts to connect to Internet with single IP 
address assigned by Internet Service Provider (ISP).
    In case you would want some traffic to reach your PC through a 
router/firewall device, you will have to configure a rule and allow it 
on router/firewall device.
2. Your PC is connected to a router device that works as a network 
bridge and your PC has public IP address assigned by ISP.
    In this case any unsolicited Internet traffic (automated 
connections, port scans, etc) will reach your PC and should be stopped 
by a firewall.
3. Your PC is connected to Internet cord directly and PC has public IP 
address assigned by ISP.
    In this case any unsolicited Internet traffic (automated 
connections, port scans, etc) will reach your PC and should be stopped 
by a firewall.


In cases 2 and 3 you have to keep firewall up and configured to block 
incoming traffic. Also you have to be aware of any active network 
services on your PC that could be accessed from the Internet and it is 
your job to keep them secure.
These services could be anything: SSH server, FTP server, HTTP server, 
SQL server, SAMBA server, game servers, etc.


In case 1 you are relatively safe from Internet traffic noise. Hosts on 
your LAN are separated from the Internet by router/firewall device.


Now, I don't want to scaremonger and feed anyone's paranoia, but for the 
sake of completion, there are known cases in history when 
router/firewall had vulnerabilities, or firmware flaws, or configuration 
negligence, that allowed perpetrators to 'hack' them, as in gain full 
access and control over their firmware and gain network access to LAN hosts.
These cases are extremely rare nowadays and very hard to pull off 
successfully, especially if the device owner keeps firmware up-to-date 
and configuration tidy.


I hope this helps you to understand a little more how networking works 
under the hood and while there is indeed a network traffic noise 
reaching every second every host on the Internet, 99.99% of it simply 
dropped by firewalls, ISP filters, or fail otherwise.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Root password strength

[Alexander V. Makartsev]

On 20.03.2024 20:28, Jan Krapivin wrote:

I must mention that "32 characters" is only my guess.

In the Handbook it is said: "The root user's password should be long 
(12 characters or more) and impossible to guess."


Also, i must again say that in my case we speak just about a humble 
home desktop, without a ""ssh" access"" or whatever complicated.


Thank you for your answers and tips. I will make a very strong 
password for root and a strong one for  a user in the sudo group.

This conclusion seems less than optimal to me.
By condemning yourself to type 12+ character password every time you 
'sudo' would really hurt accessibility and usability of your home 
computer and for no good reason.


If we focus solely on your use case: a login security of a PC at home, 
without remote access, then password of your sudo user could be as short and
simple as four numbers, of course unrelated to your date of birth, phone 
number, or any other easily guessable sequence of numbers, like '1234'.
And to prevent guessing password by "bruteforce" you will need to 
restrict number of allowed login attempts.
This could be done by enabling and configuring PAM module. ( man 
pam_faillock )
If configured correctly after a few failed login attempts user will be 
locked out for a configured amount of time and will be unlocked 
automatically once time passes.


Also think about this scenario: a visitor or relative will get physical 
access to your PC and will be able to type on keyboard, reboot it, 
access USB ports, etc.
If perpetrator could do all that, long passwords won't save you, because 
it is easy to reset passwords or add a new sudo user without knowing any 
passwords.
This could be done by simply booting to live OS on USB drive and 
'chroot' into filesystem of your OS.


To defend from this scenario you need to have encrypted filesystem with 
a strong password and never leave your PC with logged in session.
Logged in user session could be used by hackers in theory and practice 
to exploit a known (unpatched) or an unknown (0-day) vulnerability and 
escalate user privileges.

Of course, these hackers have to come into your house first. :)


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: testing new sdm drive

[Alexander V. Makartsev]

On 10.02.2024 03:34, gene heskett wrote:

On 2/8/24 07:22, Alexander V. Makartsev wrote:

This is how I would test it.
First create a new GPT partition table and a new 2TB partition:
 $ sudo gdisk /dev/sdX check

/!\  Make double sure you've selected the right device by using 
"lsblk" and "blkid" utilities.  /!\
/!\    It could change from 'sdm' to another 
name after reboot.      /!\


At gdisk prompt press "o" to create a new GPT table, next press "n" 
to create a new partition, accept default values by pressing "enter".
To verify setup press "p", to accept configuration and write it to 
device press "w". check


Next format partition to ext4 filesystem:
 $ sudo mkfs.ext4 -m 0 -e remount-ro /dev/sdX1 check

Next mount the filesystem:
 $ sudo mkdir /mnt/disktest check
 $ sudo mount /dev/sdX1 /mnt/disktest check

Next create reference 1GB file filled with dummy data:
 $ cd /mnt/disktest check
 $ sudo fallocate -l 1G ./reftestfile check
 $ sudo badblocks -w -s -t random ./reftestfile check

Now we can use script to create 1830 1GB files and check their checksum:
 $ for i in $(seq 1830); do sudo dd if="./reftestfile" 
of="./testfile${i}" status=none; md5sum -b "./testfile${i}" ;done


This procedure will take a very long time to complete. "md5sum" will 
output the checksum for each file and they should be equal to 
checksum of "reftestfile":

 $ md5sum -b ./reftestfile

Got a problem Alexander:
I had to put the script someplace else. So I put it in my private 
/home/gene/bin as disktest.txt with nano. couldn't find it.

But:
gene@coyote:/mnt/disktest$ sudo /home/gene/bin/disktest.txt
sudo: /home/gene/bin/disktest.txt: command not found
If you put that 'for' loop one-liner inside, I think you forgot to make 
"/home/gene/bin/disktest.txt" executable:

    $ chmod +x /home/gene/bin/disktest.txt


And:
gene@coyote:/mnt/disktest$ ls /home/gene/bin/disktest.txt
/home/gene/bin/disktest.txt
So I think I found the problem with my script, ancient eyeballs can't 
tell the diff between () and{} so I fixed that but it still won't run 
or be killed. I don't care how big you've made the t-bird font, by the 
time you've read 2 more msgs, its back to about 6 point text.  Grrr.


So I fired up a root session of htop, found about 8 copies of dd 
showing and started killing them but cannot kill the last 2 in the D 
state.


And cannot find .disktest.txt running in a root htop and the2 copy's 
of dd can't be killall'd.



It's not possible for me to know what went wrong.
Have you created "reftestfile" inside "/mnt/disktest" directory?
How many "testfile*" files, if any, were created on the filesystem 
mounted at "/mnt/disktest"?

Was there anything relevant in the syslog about "sdm" drive after the test?
If you'd followed my instructions step by step, you'd end up inside 
"/mnt/disktest" directory and for the last step all you had to do is 
copy and paste that one-liner 'for' loop into the command line.
It's a long line and it really meant to be copied and pasted not typed 
by hand, and also to give you the idea of the process, so you could 
adjust it if needed.
I've tested it again on my computer and it worked as expected, 
synchronously created "testfiles" inside current directory and 
calculated their hashes one by one.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: testing new sdm drive

[Alexander V. Makartsev]

On 09.02.2024 00:23, gene heskett wrote:
Looks neat. Any chance this will crash my machine? I have other design 
work going on, and I'd hate to have to start from scratch.
Well, it will consume CPU cycles for sure, at least to calculate md5 
hashes and perform I/O on the target drive and RAM.
I don't think it could crash the system, but the load could be 
significant enough to disturb your work, so
if I was in your place I'd wait until the machine is free from any work 
or load and then test the new drive.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: testing new sdm drive

[Alexander V. Makartsev]

On 08.02.2024 12:14, gene heskett wrote:

gene@coyote:/etc$ sudo smartctl --all -dscsi /dev/sdm
smartctl 7.3 2022-02-28 r5338 [x86_64-linux-6.1.0-17-rt-amd64] (local 
build)
Copyright (C) 2002-22, Bruce Allen, Christian Franke, 
www.smartmontools.org


=== START OF INFORMATION SECTION ===
Vendor:
Product:  SSD 3.0
Revision: 2.00
Compliance:   SPC-2
User Capacity:    2,097,152,000,000 bytes [2.09 TB]
Logical block size:   512 bytes
scsiModePageOffset: response length too short, resp_len=4 offset=4 
bd_len=0
scsiModePageOffset: response length too short, resp_len=4 offset=4 
bd_len=0

>> Terminate command early due to bad response to IEC mode page
A mandatory SMART command failed: exiting. To continue, add one or 
more '-T permissive' options.

gene@coyote:/etc$

And then again, it worked, sorta

Cheers, Gene Heskett, CET.


Looks like a scam. Probably a reprogrammed controller to falsely report 
2TB of space to the system.


This is how I would test it.
First create a new GPT partition table and a new 2TB partition:
    $ sudo gdisk /dev/sdX

/!\  Make double sure you've selected the right device by using "lsblk" 
and "blkid" utilities.  /!\
/!\    It could change from 'sdm' to another 
name after reboot.      /!\


At gdisk prompt press "o" to create a new GPT table, next press "n" to 
create a new partition, accept default values by pressing "enter".
To verify setup press "p", to accept configuration and write it to 
device press "w".


Next format partition to ext4 filesystem:
    $ sudo mkfs.ext4 -m 0 -e remount-ro /dev/sdX1

Next mount the filesystem:
    $ sudo mkdir /mnt/disktest
    $ sudo mount /dev/sdX1 /mnt/disktest

Next create reference 1GB file filled with dummy data:
    $ cd /mnt/disktest
    $ sudo fallocate -l 1G ./reftestfile
    $ sudo badblocks -w -s -t random ./reftestfile

Now we can use script to create 1830 1GB files and check their checksum:
    $ for i in $(seq 1830); do sudo dd if="./reftestfile" 
of="./testfile${i}" status=none; md5sum -b "./testfile${i}" ;done


This procedure will take a very long time to complete. "md5sum" will 
output the checksum for each file and they should be equal to checksum 
of "reftestfile":

    $ md5sum -b ./reftestfile
    3f2c5fa95492bfaa18f08c801037d80b *./reftestfile

    $ for i in $(seq 1830); do sudo dd if="./reftestfile" 
of="./testfile${i}" status=none; md5sum -b "./testfile${i}" ;done

    3f2c5fa95492bfaa18f08c801037d80b *./testfile1
    3f2c5fa95492bfaa18f08c801037d80b *./testfile2
    ...
    3f2c5fa95492bfaa18f08c801037d80b *./testfile1830

Obviously, checksum for your "reftestfile" will be different from mine.
If 'for' loop fails at some point, you can count testfiles to see how 
many of them were actually written to disk.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Mariadb error

[Alexander V. Makartsev]

On 19.01.2024 13:46, Gift Rain wrote:

Good day,

I'm running Debian 11 and getting error "Failed to start MariaDB 
10.5.12 database server.


Your MySQL database server doesn't start for some reason. Is there 
anything helpful in the output?


   $ sudo journalctl --no-pager -x --unit mariadb.service

You can use Debian "pastebin" site [1] to send the output if it is long.

What happened before failure? Was it system crash? System upgrade 
attempt? Anything else?




My emails are not working. Another error on my logs is :

"warning:proxy:mysql:/etc/postfix/mysql/sender_bcc_maps_user.cf 
 lookup error for"


Postfix uses mysql database to lookup necessary information and since it 
is not running so does postfix.



[1] https://paste.debian.net/
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: 512e vs 4K sector confusion

[Alexander V. Makartsev]

On 14.01.2024 13:15, Andy Smith wrote:

On Sun, Jan 14, 2024 at 08:01:52AM +, Andy Smith wrote:

If necessary and if there is a way, I *can* nuke off the target
machine's "foo" volume group and recreate the RAID array if I have
to make it 512e format. But obviously I'd like some way to move this
disk image and have it still work without having to meddle inside it
much — it is a VM disk.

I think I may be able to use hdparm to reformat these Ultrastar DCs
from 4kn to 512e…

hdparm --set-sector-size 512 --please-destroy-my-drive /dev/sdX

https://wiki.archlinux.org/title/Advanced_Format#Advanced_Format_hard_disk_drives

It is okay to destroy the drive contents as there's nothing working
on there yet given this initial failure.

I think that is the only way this is going to work without mounting
filesystems both sides and using an fs-aware tool like tar|ssh or
rsync… which I don't want to do.

Thoughts?

Weird. I can imagine some wild theories. :)
I think it has something to do with a fact that "disk_image" is a LVM 
volume and you dd copy "/dev/mapper/VG--name-LV--name" not "/dev/sdc2" .
Could it be that "dd" sees source and destination as block devices and 
copies data by sector? I.e. takes 512b sector and place it to the 
beginning of 4096b sector and padding the rest 3584b of the sector with 
zeroes.
This calls for some tests. I'd create 1G image test file with dos 
partition table and a few primary partitions:


   $ dd if=/dev/zero of=./1G-disk-image.bin bs=1M count=1000
   1000+0 records in
   1000+0 records out
   1048576000 bytes (1,0 GB, 1000 MiB) copied, 8,07768 s, 130 MB/s

   ...
   $ /sbin/fdisk -l ./1G-disk-image.bin
   Disk ./1G-disk-image.bin: 1000 MiB, 1048576000 bytes, 2048000 sectors
   Units: sectors of 1 * 512 = 512 bytes
   Sector size (logical/physical): 512 bytes / 512 bytes
   I/O size (minimum/optimal): 512 bytes / 512 bytes
   Disklabel type: dos
   Disk identifier: 0x143acb59

   Device   Boot   Start End Sectors  Size Id Type
   ./1G-disk-image.bin1 2048  411647  409600  200M 83 Linux
   ./1G-disk-image.bin2   411648 1435647 1024000  500M 83 Linux
   ./1G-disk-image.bin3  1435648 2047999  612352  299M 83 Linux


And dd copied it the same way to disk with 4k sectors.
Just to see if it will be inconsistent too, without LVM mapper involved.

I'd also try to clone disk to image file first, with say "partclone" 
utility, "scp" the image file and restore it at remote destination with 
the same utility after recreating dumped partition table layout using 
"sfdisk".
I did this procedure multiple times successfully, but without 4k sector 
disks and LVM involvement.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

OT: Pitfalls of online purchases (Was Re: Mouse single click handling?)

[Alexander V. Makartsev]

On 19.12.2023 23:48, Felix Miata wrote:


Can you suggest any particular online source in North America that sells those
switches? I had no success trying to refurb one in my invaluable Logitech 
Trackman
Marble FX PS/2 trackball. It took 2 hours to get that tiny switch back together
after disassembly and cleaning. :( A Kensington Orbit I had much better luck 
with
cleaning, so didn't need a new switch.

I can't unfortunately.


AliExpress is reputable? That's news to me. I ordered something from it 20 
months
ago. There was a lot of back and forth email between us before I actually got
something a month later that turned out to be used rather than new, and just as
broken as the original part I was trying to replace. My credit card company
eventually issued a refund after I explained all that happened. That particular
card I rarely used, and hadn't for a number of months before that online 
purchase
or for several months after. 3 months or so after the refund, there were 4 cash
advances issued on it totaling around $8,000 on the other side of an ocean I had
never crossed. Those got reversed, but my credit rating dropped 100 points and
hasn't come near recovering.

I share information only about my own experiences with AliExpress.
It might be very hard to find a good and reliable store there, but at 
least their official support and "dispute-system" are working fine.
The important part is that you have to keep all transactions, 
communications, chats within AliExpress platform.
Never agree to switch to email or any other resource for conversations 
with sellers. It's just asking to be scammed and AliExpress official 
customer service will be unable to help you.
I had to open disputes a few times and usually AliExpress officials 
step-in and resolve the dispute in customer's favor, unless the dispute 
is unreasonable and there is no evidence.
The whole shopping experience there could be daunting, there is simply 
too much stores and wares to choose from and their site is not helping, 
by being difficult and slow.
IMO it is too much of a risk to buy anything expensive there, given that 
postal service might smash the package during delivery, but ordering 
something like electronic components, hand tools, repair equipment, 
spare parts, etc, should be fine.
I had to order spare LCD screens and fragile repair equipment from 
reputable shops a few times. All delivered without damages.
Sometimes shops send items with a wrong part number, I had to start 
disputes explaining the problems and return items back. All returns were 
refunded back to my debit card.
I've never encountered any problems with payments or refunds, and their 
Escrow-account protection ensures that money will be released to the 
seller only after customer has received what they ordered.
I've heard there is a frequent "carder" and "skimmer" problem in the 
United States [1], so those issued card advances you had might be 
nothing to do with AliExpress, if you paid using their platform.
Anyway, I've used many cards throughout the years to pay for goods on 
AliExpress and never encountered any suspicious activities.
I also extra careful when I have to reveal my card information online 
and when I use ATMs.



[1] https://krebsonsecurity.com/?s=carder+skimmer
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Mouse single click handling?

[Alexander V. Makartsev]

On 19.12.2023 21:41, to...@tuxteam.de wrote:

On Tue, Dec 19, 2023 at 09:37:20PM +0500, Alexander V. Makartsev wrote:

On 19.12.2023 20:04, local10 wrote:

Hi,

I have several mice that went bad with the same defect: they sometimes generate 
two single clicks very quickly (say, within 10-20ms) instead of a single click.

This is a very common problem with micro-switches inside computer mice,
mechanical keyboards, etc.
They wear out with use and has to be replaced, given that they are very
inexpensive, standardized and relatively easy to replace.

I've had some success inserting a thin shim between plunger and
switch, on the theory that the whole thing (plunger, switch, mount
on the PCB) get shortened/worn out on use.

Usually it is gunk and taint getting accumulated on metal contacts inside.
If you really have to repair a micro-switch without soldering, you can 
disassemble them, clean thoroughly contacts inside and metal plate with 
Isopropyl alcohol, let it dry out and reassemble.

Required tools:
1. Needle to pry open a plastic cap and tweezers (those could be bought 
at hardware store)

2. Steady hands (can't be bought)
3. Good eyesight (you guessed it, also can't be bought).


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Mouse single click handling?

[Alexander V. Makartsev]

On 19.12.2023 20:04, local10 wrote:

Hi,

I have several mice that went bad with the same defect: they sometimes generate 
two single clicks very quickly (say, within 10-20ms) instead of a single click.
This is a very common problem with micro-switches inside computer mice, 
mechanical keyboards, etc.
They wear out with use and has to be replaced, given that they are very 
inexpensive, standardized and relatively easy to replace.
Here is a link to PDF datasheet [1] for Omron D2F series switches 
specifications.
You need a switch with pin plunger and compatible terminals, those 
depend on mice model.
You can buy them at local electronics store, order them from China 
(AliExpress), or from any other reputable source.



[1] https://omronfs.omron.com/en_US/ecb/products/pdf/en-d2f.pdf
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Problem with Xfce4 Applications Menu

[Alexander V. Makartsev]

On 13.12.2023 00:40, Stephen P. Molnar wrote:


I have just bumbled my way thorough reinstalling v 12.4.0 on my main 
Linux platform and have managed to mangle the Xfce4Applications Menu 
by somehow misusing MenuLibre.


Some how the contents of some of the subdirectories have been shifted 
to an 'Other' subdirectory. My question is how do i reconstruct the 
Applications Directory?


I don't know anything about MenuLibre, but maybe Xfce Application Menu 
was switched to another menu file?

Have you tried to switch it back to default menu?
RMB on Application Menu icon --> Properties --> Use the default menu


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Boot Problem

[Alexander V. Makartsev]

On 05.12.2023 23:33, Stephen P. Molnar wrote:
I decided to try something. I logged in to the rescue  mode as root 
and entered startx at the prompt. This generated the error:


Unable to contact settings server
failed to execute child process "dbus-launch" (No such file or directory)

I vaguely remember having similar behavior as you described in the first 
message, but I can't recall what was the root cause of that problem.
Probably it was Xfce4 user session failing to start because of config 
file or cache corruption.
Try to create a new test user and login using these new credentials, 
just to test it out.


It would be great if you'd also share an information about your system, 
installed version of Xfce, etc.
Contents of log files in your user's home directory "~/.xsession-errors" 
and "~/.xfce4-session.verbose-log".

You can use Debian Pastezone [1] to share long text files.


[1] https://paste.debian.net/
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Running 32 bit apps on 64 bit debian

[Alexander V. Makartsev]

On 19.11.2023 12:58, Van Snyder wrote:

I'm trying run 32 bit LinuxSusser on 64 bit Debian 12 bookworm.

When I try to run it, I get
./LinuxSusser: Command not found.

"ls -l ./LinuxSusser" respnds
-rwxr-xr-x 1 vsnyder vsnyder 12698092 Feb  8  2013  LinuxSusser*

"dpkg --print-architecture" responds amd64
"dpkg --print-foreign-architectures" responds i386

"file ./LinuxSusser" responds
./LinuxSusser:  ELF 32-bit LSB executable, Intel 80386, version 1
(SYSV), dynamically linked, interpreter /lib/ld-linux.so.2, for
GNU/Linux 2.6.0, stripped

but "ldd ./LinuxSusser" responds "not a dynamic executable"

I used to run it in Debian 10.

What am I doing wrong?


It starts on my "bookworm" machine.
Your executable probably got corrupted somehow.
Try to download it again from author's website [1]. There you will find 
a link to DropBox file sharing service.

Install dependencies (64bit and 32bit) for this program:

   $ sudo apt install libglib2.0-0 libpango-1.0-0 libatk1.0-0 openssl
   zlib1g libexpat1

   $ sudo apt install libglib2.0-0:i386 libpango-1.0-0:i386
   libatk1.0-0:i386 zlib1g:i386 libexpat1:i386

A normal output from "ldd" and "file" for reference [2].


[1] https://www.madoverlord.com/wiki/doku.php/madoverlord:projects-sudoku
[2] https://paste.debian.net/plain/1298552
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Hardware Advice Wanted: Router

[Alexander V. Makartsev]

On 16.11.2023 03:46, Charles Curley wrote:

On Thu, 16 Nov 2023 01:58:05 +0500
"Alexander V. Makartsev"  wrote:


16 years is a good amount of value. :)
Is it Pentium 4 on ITX motherboard?

Nope. FIT-PC, first iteration. Processor is an AMD Geode SBC.
https://linux-hardware.org/?probe=c256a73072


https://www.compulab.com/products/computer-on-modules/cm-iglx/#specs
x86 CPU 500Mhz, DDR1, IDE ATA-100, 100Mbit NIC and only 3-5W.
Cool stuff.
I've never seen them in the wild.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Hardware Advice Wanted: Router

[Alexander V. Makartsev]

On 15.11.2023 18:47, Charles Curley wrote:

On Wed, 15 Nov 2023 12:31:52 +0500
"Alexander V. Makartsev"  wrote:


On 15.11.2023 07:56, Stefan Monnier wrote:
  [...]
  [...]
I wrote that email as a word of caution, because Roberto had
mentioned he is looking for the device with the same conditions as
OP, which is "no fans".
And this model will be very noisy at all times.

How do you figure that it will be noisy at all times? I saw noting in
the specifications or reviews that indicated that.

Are you talking about Neosmay based on i7-1260P CPU?
Like I said previously there are no reviews that would show temperatures 
under load and no disassembly videos that will show the thermal solution.
IMO the reason for that is simple, to keep it quiet about inconvenient 
technicalities and to sell sub-optimal product.
Specifications for this CPU shows a requirement of a good thermal 
solution, which should be adequately large\noisy for this form factor.



  [...]
And this is why there is no reason to have high power CPU inside.
I expect the CPU temperature to be at 98 degrees Celsius at all times
and constant throttling under minimal load.

Yikes! None of my machines gets anywhere near that. The fan cooled CPUs
on my desktop are in the 26-32° range right now (early morning, no
load). The machine I am typing this on, a laptop, is reporting core
temps of 44° and 50° right now. All consider high to be ~~86°, with
critical near 98°. (Data courtesy of the sensors program.)
None of my personal machines either. It all depends on specifications 
for each individual piece of hardware.
My desktop has Skylake CPU rated at 65W Base TDP and cooled by 
tower-style heatsink with 4 heatpipes rated at 130W TDP.
Idle, small load temps are 28-32°C, busy workloads: 4 cores compiling, 
7-zip compression, video rendering, maxes out at 70°C in Summer.
There is also a netbook, I use it on rare occasion, it has 6.5W Atom CPU 
and Nvidia ION discrete VGA.
It has heatsinks for both Intel SoC and Nvidia chipset with heatpipes 
and a cooling fan, and it is incapable of overheating even under high 
load, because of low power CPU.
This netbook is painfully slow, it has quite noisy small fan, but 
temperatures are maxed out at 50°C.
The thing is desktops and laptops could have much better thermal 
solution simply because there is space inside for it.


Now let's look at thermal solutions of Intel NUCs [1] [2] for an example.
They look similar to what you could see inside a laptop, except in both 
cases there is only a small heatsink and two heatpipes.

VRM zone is cooled by air which is not great.
First photo depicts thermal solution for i5-10210U processor [3] which 
is rated at 15W TDP, or 10-25W base frequency.
Second photo depicts thermal solution for i7-1165G7 processor [4] which 
is rated at 12-28W base frequency.
Both thermal solutions show sub-par performance under load, because 
there are multiple complains about overheating NUCs on the Internet.
Most of the solutions for the complaints either a cleanup from dust or 
set a power limit via BIOS, or both.
There is a fan-less case [5] for NUCs is available, which looks 
reasonable, basically a brick of aluminum, but even then they are rated 
25W TDP maximum and

it is required from user to set BIOS settings to power limit CPU to 25W.

And that's about widely distributed and supported NUCs from Intel.
Now lets take a look at Newsmay Neosmay products. Odd naming.
We can't find any photos of disassembled products on Newsmay official 
website [6].
There is one good review of S2-B560TPM mini PC [7] with photos of it 
disassembled.

And there it is, same two heatpipes and a heatsink with a fan.
As shown in the review, even with slightly larger case size, they 
couldn't manage to design adequate cooling solution for i5-11400T 
desktop CPU [8] which is rated 35W TDP,

although there are additional heatsinks covering VRM zone, which is a plus.
It looks like a good thermal solution for a 10-15W TDP CPU and that is it.

So once again, what is the point to have high power CPU inside (and pay 
more for it also),
if because of inadequate thermal solutions you have to power limit it to 
base frequency, or suffer from constant throttling under load, noise and

other consequences of overheating, such as reduced lifetime of the device?


What good will it be if with high probability the device will burn
out in 3 months?

Indeed. I've gotten 16 years out of my FIT-PCs so far and would like to
get a respectable portion of that out of their replacement.

16 years is a good amount of value. :)
Is it Pentium 4 on ITX motherboard?
Nowadays they don't make them like before, and it is so hard to buy 
something decent.
Everything is power hungry, working at insane frequency speeds; tiny ICs 
are more fragile and susceptible to overheating;
BGA SoCs with hundreds of leads on thin PCBs prone to deformation; 
lead-free solder and all that other planned obsolescence dance..



[1] 
https://nucblog.ne

Re: Hardware Advice Wanted: Router

[Alexander V. Makartsev]

On 15.11.2023 07:56, Stefan Monnier wrote:

This looks too good to be true and raises many red flags.
According to Intel specs [1] for this processor it's 28W of heat to
  dissipate and that is Base Power only, Turbo Boost is whooping 64W(!).
IMO it is impossible to do with fan-less design at this small size, so there
will be at least small fan running at full speed at all time and there has
to be quite beefy heatsink also.

The marketing material does say fairly clearly that there's a fan inside
(it even boasts of it being "efficient" and "low noise", mentioning
"45db").
I wrote that email as a word of caution, because Roberto had mentioned 
he is looking for the device with the same conditions as OP, which is 
"no fans".

And this model will be very noisy at all times.


Also my understanding is that this CPU is like the one in many NUCs which
means that its 4.7GHz turbo mode can be used only for *very* short
bursts, so it may be unnecessary to size the heatsink for the full
64W as long as it can withstand it for a long enough "short burst".

And this is why there is no reason to have high power CPU inside.
I expect the CPU temperature to be at 98 degrees Celsius at all times 
and constant throttling under minimal load.
There is also VRM zone often neglected, most likely it will be air 
cooled (without heatsink) with temperatures close to 100C at all times and

there is a limit of how much heat a PCB could dissipate until carbonation.
Even NVMe SSD and RAM generate substantial amounts of heat and I doubt 
this device has the same design as NUC that has extra heatsink under the 
top cover.
What good will it be if with high probability the device will burn out 
in 3 months?



Note they also offer a fanless system with somewhat similar specs for
$100 more.  That sounds a lot more atttractive to my ears.
Intel NUC with the same CPU [1] and fan cooling, draws 40-90W from PSU 
and struggle with throttling, overheating and noise.
And NUCs are usually expected to have much better thermal solution 
design than its clones.


If I needed a fan-less PC of this form factor I'd look for 10-15W low 
power CPUs and
refused to buy anything if there is no video review of it, with complete 
disassembly and temperatures under load,

or at least a complete service manual with accurate pictures of components.


[1] https://www.youtube.com/watch?v=fZbYSUwvea0
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Hardware Advice Wanted: Router

[Alexander V. Makartsev]

On 14.11.2023 04:08, Roberto C. Sánchez wrote:

On Mon, Nov 13, 2023 at 03:57:28PM -0700, Charles Curley wrote:

My FIT-PCs that provide network services are getting old, and i386
Linux is slowly fading away. So I would like to replace them with a
router/gateway computer.

It should run Debian.

It should either have two gigabit (or better) Ethernet interfaces or
have suitable expansion capability.

It should be quiet: no fans, and low power requirements. A small
physical footprint would be nice.

Most of the time it will run headless, but occasionally I will need to
stick a monitor and keyboard on it. VGA will do fine.

It will be a router. It will have at least DNS, DHCP, apt-cacher-ng and
firewalld on it. Modest disk and RAM will be fine.

WiFi is handled elsewhere for now, but I won't turn it down.

SSD for storage would be nice, although the FIT-PCs do fine with IDE
spinning rust.


I am in the process of solving almost the exact same problem. At the
moment, this is my leading candidate:
https://www.newegg.com/neosmay-kc12-alder-lake/p/2SW-006Y-00074

This looks too good to be true and raises many red flags.
According to Intel specs [1] for this processor it's 28W of heat to 
dissipate and that is Base Power only, Turbo Boost is whooping 64W(!).
IMO it is impossible to do with fan-less design at this small size, so 
there will be at least small fan running at full speed at all time and 
there has to be quite beefy heatsink also.
There are no photos of this device disassembled and fake 3D rendered 
pictures doesn't count.
A quick search of "NEOSMAY MINI PC Disassembly" on YT gave no results, 
only fake reviews and useless unboxing videos.
I've seen 17" laptops with much larger heatsink solutions (multiple 
heatpipes and fans), struggle with effective heat dissipation.



[1] 
https://www.intel.com/content/www/us/en/products/sku/226254/intel-core-i71260p-processor-18m-cache-up-to-4-70-ghz/specifications.html

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Amazing how far things have come. a 32x100G switch running Debian.

[Alexander V. Makartsev]

On 12.11.2023 23:34, Andy Smith wrote:

On Sun, Nov 12, 2023 at 05:48:27PM +, Andy Smith wrote:

Well done Mellanox, and Debian. I hope to see more of it!

…although I did forget that Nvidia acquired Mellanox in 2019 and
since then has scrapped the Mellanox brand name, so the good times
are probably over. 

Yeah, tech industry always kills companies who wanted to make something 
great.
Gotta keep spoon-feeding the technology and god forbid the big tech 
makes something too good, or too advanced, or too robust.


Nice article. I felt quite skeptical about "two core Celeron CPU" until 
I got to ASIC part. :)
I've never had a chance to build something more advanced than a basic 
network router with 2 ethernet interfaces. VM-based toys don't count. :)

Still it was interesting to learn about switchdev.
Now I wonder could it be used for non-Mellanox hardware..


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to compare one folder to one directory

[Alexander V. Makartsev]

On 03.11.2023 08:17, Greg Wooledge wrote:

Calling these things "folders" discards all of this history and
knowledge.

But the real problem with calling them "folders" is that it doesn't
match the Unix user interface.
Personally, I don't see the problem, because I was talking to people not 
Unix user interfaces.
I can agree that word "directory" is more correct scientifically and 
word "folder" sound simplified, but
so does word "hammer" sound more simple than scientifically correct 
"hand tool for nail punching".
In the end both names mean the same thing and people will always 
interpret them correctly, which is the most important part.

And if I had to talk to Unix user interface I'd speak in system calls and
expected neither "directories" nor "folders" as answers, but data 
structures and return codes instead.


How is this still going?

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to compare contents of two folders against third one?

[Alexander V. Makartsev]

On 02.11.2023 12:59, to...@tuxteam.de wrote:

Rsync wasn't helpful in my case. It refused to recognize file names
properly, complaining about charset encoding, not even with --iconv
parameters.
Probably because of locale differences between my system and mounted
filesystems and who knows what else.

Difficult to say without knowing more details. Are those "directories"
actually on different file systems?

For vfat, specifically, there is a codepage argument for mount.

Cheers
I've tried to mount filesystems (all NTFS) with different locale 
parameters, still ended up with either garbled filenames or charset 
conversion and 'path too long' complains from rsync.
Since I've copied files manually with Thunar, it didn't complained about 
anything and all copied files were consistent, even those rsync 
complained about.
It could be also a limitation or bug of overlayfs since it doesn't have 
locale/iocharset/codepage parameters for mount.
Anyway, I don't have access to source and two destination disks anymore 
to investigate this further.

Data recovery was successful and case closed.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to compare contents of two folders against third one?

[Alexander V. Makartsev]

On 01.11.2023 22:55, to...@tuxteam.de wrote:

I concur with Nicolas: every time you say "folder", a unicorn dies.


If I say tater instead of potato would it also make you sad?
Sorry to break it to you, but unicorns were extinct a long time ago, 
along with BBS', modems, MSDOS, ISA-VLB slots, jumpers and through-hole ICs.
Only BGAs, UEFI, liquid metal, Bluetooth and other planned obsolescence 
garbage remains.



You mean: the union of dest-dir-1 and dest-dir-2 should equal your
source-dir?

   Try rsync -a source-dir/ dest-dir-2/ --compare-dest=dest-dir-1/

This one will even dutifully copy those files from source-dir which
can't be found either in dest-dir-1 or in dest-dir-2.

If you only want to /see/ what would be copied (always a good idea
when trying untested advice from random folks on the internets :)
there's the option --dry-run. Perhaps add the option -v.

Rsync is magic.

Cheers
Rsync wasn't helpful in my case. It refused to recognize file names 
properly, complaining about charset encoding, not even with --iconv 
parameters.
Probably because of locale differences between my system and mounted 
filesystems and who knows what else.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to compare contents of two folders against third one?

[Alexander V. Makartsev]

On 01.11.2023 20:07, Nicolas George wrote:

Hi.

Alexander V. Makartsev (12023-11-01):

I have a "/source-folder/" which contains very large tree of folders and
files.

The word is “directory”, not “folder”.


Oh, I'm so sorry, I didn't meant to invoke wrath of the ancients.
I thought about it and eventually decided to name sample directories - 
folders.
Mostly because I didn't wanted to answer possible questions like "wtf is 
a directory?".
And I must to warn you, if you open Xfce file manager (Thunar), that 
blasphemous word "folder" will be everywhere!



Now, is there an effective way to compare combined contents of two folders
"/destination-folder-one/" and
"/destination-folder-two/" against a "/source-folder/" to show if there is
anything that was left out?

When you say “combined content”, you mean merged, right? There are a lot
of different ways to combine.

Yes, that is exactly what I meant.


You could mount -t overlay your two destination directories together and
try rsync or diff.


I've never heard about "mount -t overlay" before and it did the trick.
Thanks for suggestion.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

How to compare contents of two folders against third one?

[Alexander V. Makartsev]

Hello everyone.

I have a "/source-folder/" which contains very large tree of folders and 
files.
I've manually copied a set of folders and files from it to a 
"/destination-folder-one/" and

copied another set of folders and files to a "/destination-folder-two/".

Now, is there an effective way to compare combined contents of two 
folders "/destination-folder-one/" and
"/destination-folder-two/" against a "/source-folder/" to show if there 
is anything that was left out?


For now I've tried "diff" and "rsync" to accomplish this.
But diff apparently can't compare combined contents of two folders with 
another folder.


   $ diff -r /destination-folder-one/ /destination-folder-two/
   /source-folder/
   diff: extra operand '/source-folder/'
   diff: Try 'diff --help' for more information.


And rsync scans folders and files that are already exist inside both 
destination folders and "/source-folder/" and outputs nothing.


   $ rsync -r --size-only --dry-run /destination-folder-one/
   /destination-folder-two/ /source-folder/
   $


I could go on a wild chase to "ls" contents of both destination folders, 
concatenate the results, sort them somehow,
do the same to a source folder and compare the resulting list files. 
Also create test cases to check if results are reliable.

But before I do that, is there a better way to accomplish the task?
Maybe some parameter for diff or rsync that I missed or another utility?

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: EASY way to install packages from trixie/sid to stable?

[Alexander V. Makartsev]

On 26.10.2023 18:39, Hans wrote:

Hi folks,

is there a very easy way, if I want to install packages from trixie oder sid
into my bookworm installation?

I read about apt pinning, but as far as I understood, I have to name
explicitily each package I want to install from sid. This can be much work,
when installing a high number of packages.

I suppose, I then have also to install all dependencies of the packaes from
sid, even if they are related to the system.

In my case I wanted to install virtualbox from sid, as it has all packages
ready. However, virtualbox requires and depends also the newer gcc compiler
and some compiler libs, thus I took distance from installing for now.

At the moment I am not using pinning. My actual way of doing is

1. adding the sid repo into /etc/apt/sources.list

2. then aptitude -u

3. then searching for the required package and mark it as install (or
upgradeble

4. Then install, if wanted.

Yes, I know, pinning would be the better way, but it is very, very seldom, I
need to install something from a higher repo.

And yes, I know, mixing repos is no good idea, so I am using this only for
applications, which are using theire own libraries (or libs, they are only for
this special application).

Do you know another way, except pinning or my (weired) way?
I've always did it "the right way", by making simple backports [1] of 
required package and its selected dependencies if newer versions are 
required.
For some exotic packages this approach is not feasible, because you 
might end up with dozens of packages to backport as dependencies and 
dependencies of their dependencies.

"php", "nodejs" and "golang" to name a few tend to snowball a lot.

In your situation, it could be better to use officially distributed 
package [2] from Oracle for now.
Foreign packages like these could be installed into separate location 
like into "/opt/". This way they won't interfere with the rest of your 
system.
I don't use virtualbox (KVM does everything and more for me) so I can't 
vouch for the quality of packages from Oracle.


[1] https://wiki.debian.org/SimpleBackportCreation
[2] https://www.virtualbox.org/wiki/Linux_Downloads

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Hidden UUID?

[Alexander V. Makartsev]

On 18.10.2023 12:01, Hans wrote:

Hi folks,

maybe you can help. I am running into a strange problem.

It is the following:

My swap partition is a logical partition on an extended partition.
This swap was formerly luks encrypted and got an UUID beginning with
UUID=30e885.

Then I deciced, to format the swap partitition and use it as a normal swap
partition. So I erased all keys (using cryptsetup erase /dev/sda5) and then
reformatted it with swapon.

Additionally I removed the entry from /etc/crypttab and edited /etc/fstab.

So far, everything worked fine, except next boot.

When booting, cryptsetup wants to open my other encrypted partitions and then
it is hanging, saying "searching for encrypted device UUID=30e885".

I looked everywhere, but I found no entry, where the system gets this UUID.

Did you also changed initramfs resume config file, which designates a 
swap partition to use for Hibernation?


   $ /sbin/blkid | grep swap
   /dev/sdb6: LABEL="swap" UUID="e990fa6d-3f82-4d65-b6a1-542e240718fc"
   TYPE="swap" PARTLABEL="Debian-swap"
   PARTUUID="addbc03b-f50e-4175-95b4-91b7b1c8fdca"

   $ cat /etc/initramfs-tools/conf.d/resume
   RESUME=UUID=e990fa6d-3f82-4d65-b6a1-542e240718fc

Don't forget to update initrd file if you change resume file:

   $ sudo update-initramfs -u
   update-initramfs: Generating /boot/initrd.img-6.1.0-13-amd64


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: No wifi on debian 12 fresh install (HP laptop 440 14" G10)

[Alexander V. Makartsev]

On 04.10.2023 17:08, Itay wrote:

Hi

I recently purchased an HP ProBook 440 14" G10 laptop.
This configuration is certified by Ubuntu[1].
The specs, as advertised by HP, can be found here[2].

The installation (debian 12 'bookworm') went as expected (using DVD and wired
connection).  However, after reboot, I discovered that wifi was not working.
Bluetooth, on the other hand, is working.

After three days of search on the internet[3]-[9] I am really at a loss.
I have been a long-time debian *user* -- but am not an expert.
I am aware of the possibility to download drivers and compile them from
source[10]-[12], but hesitate to do that without an expert guide.

System and device information, as gathered by wireless-info script[13], will
follow the references.

Please advise
Thanks in advance.
Itay


...
# kernel 

Linux 6.1.0-10-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.38-1 (2023-07-14) x86_64 
unknown unknown GNU/Linux
...
# lspci #

03:00.0 Network controller [0280]: Realtek Semiconductor Co., Ltd. Device 
[10ec:c852] (rev 01)
Subsystem: Hewlett-Packard Company Device [103c:88e5]
...

https://forums.debian.net/viewtopic.php?t=155976

You need [1] to compile a kernel module for your wifi adapter. There is 
no other way.

Just follow build instructions [2] closely and you should be fine.


[1] https://forums.debian.net/viewtopic.php?t=155976
[2] 
https://github.com/lwfinger/rtw89#dkms-packaging-for-debian-and-its-derivatives


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: memtest86

[Alexander V. Makartsev]

On 14.09.2023 01:54, Tom Browder wrote:

On Wed, Sep 13, 2023 at 3:48 PM Tom Browder  wrote:

On Wed, Sep 13, 2023 at 3:32 PM Tom Browder  wrote:

On 13.09.2023 19:10, Tom Browder wrote:

Here I am again seeking help. I have used memtest86 long ago when I
I see that it's a Debian package, and  I installed it. Now I see
memtest86 on my boot choice screen, but selecting memtest86 does nothing.


That's weird.

I just tried again with only memtest86+. After reboot, I get several
memtest86+ options but when I select one I gect a pretty Debian screen
with nothing apparent happening.

Retried, nothing.

Is it possible that you forgot to disable UEFI Secure Boot feature in BIOS?
There is a FAQ section on an official website [1], that could help you 
troubleshoot boot problems.



[1] https://www.memtest.org/
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: memtest86

[Alexander V. Makartsev]

On 14.09.2023 00:42, gene heskett wrote:

On 9/13/23 12:40, Alexander V. Makartsev wrote:

On 13.09.2023 19:10, Tom Browder wrote:
Here I am again seeking help. I have used memtest86 long ago when I 
burned it on a CDROM disk.


I see that it's a Debian package, and  I installed it. Now I see 
memtest86 on my boot choice screen, but selecting memtest86 does 
nothing.



That's weird.
It works for me, although I use back-ported version 6.20 of 
"memtest86+" from Testing with a cosmetic patch. I like my grub menu 
tidy. :)
I've now noticed you've wrote "memtest86", did you meant it to be 
"memtest86+"?


memtest86 has its roots in 8086 16 bit code, and its been quite a 
party for the coders to first bring it up to 32 bit, and finally to 64 
bit. That last version I downloaded and burned was memtest86 V9.4 
which works on my 6 core i5 as well as it did on 8086's but of course 
a bit faster.
A google search should get you a link to burn to a new cd/dvd, and it 
Just Works.
I know "memtest86" was before "memtest86+", but memtest86+ is a 
successor [1] and up-to-date version should work for an old and modern 
hardware.

This is probably the reason why "memtest86" wasn't included to Stable repos:

   $ rmadison memtest86
   memtest86  | 4.3.7-3   | oldoldstable | source, amd64, i386
   memtest86  | 4.3.7-3   | oldstable    | source
   memtest86  | 4.3.7-3+b1    | oldstable    | amd64, i386

   $ rmadison memtest86+
   memtest86+ | 5.01-3 | oldoldstable   | source, amd64, i386
   memtest86+ | 5.01-3.1   | oldstable  | source, amd64, i386
   memtest86+ | 6.10-2~bpo11+1 | bullseye-backports | source, amd64, i386
   memtest86+ | 6.10-4 | stable | source, amd64, i386
   memtest86+ | 6.20-3 | testing    | source, amd64, i386
   memtest86+ | 6.20-3 | unstable   | source, amd64, i386

The other reason is that "memtest86" was sold to PassMark [2] and became 
closed source since version 4.3.
Anyway, I'd consider "memtest86+" the only reliable OSS option, which 
I've used successfully for many years with Legacy BIOS/UEFI and PXE.



[1] https://www.memtest.org/
[2] https://en.wikipedia.org/wiki/Memtest86
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: memtest86

[Alexander V. Makartsev]

On 13.09.2023 19:10, Tom Browder wrote:
Here I am again seeking help. I have used memtest86 long ago when I 
burned it on a CDROM disk.


I see that it's a Debian package, and  I installed it. Now I see 
memtest86 on my boot choice screen, but selecting memtest86 does nothing.



That's weird.
It works for me, although I use back-ported version 6.20 of "memtest86+" 
from Testing with a cosmetic patch. I like my grub menu tidy. :)
I've now noticed you've wrote "memtest86", did you meant it to be 
"memtest86+"?




I assume I probably have to add something to the grub2 menu to use it.


Nothing has to be added, it should work from the grub menu.
On what hardware did you run it?
There could be still some bugs left, and throughout the years, I've seen 
some of them, usually BIOS and/or Integrated VGA related.



I have searched for how to do that, but all I've been able to find are 
instructions on how to use it from a bootable USB thumb drive.


My grub/boot foo has disappeared. Any help on using the installed 
memtest86 is greatly appreciated.


As a fallback, I will go the USB route if I have to, :-(

USB route is usually the way to go, especially when "memtest86+" is a 
part of some recovery tools project like "System Rescue". [1]




[1] https://www.system-rescue.org/
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Nvidia 390 driver no longer available for Bookworm; nouveau constantly freezes. Solutions?

[Alexander V. Makartsev]

On 17.08.2023 20:48, Luiz Romário Santana Rios wrote:


Hello, all,

(Please cc me when replying, I'm not subscribed to the mailing list)

I have (according do lspci) a NVIDIA Corporation GK208B [GeForce GT 
710] (rev a1) graphics card. The correct proprietary driver for this 
card seems to be the Nvidia 390 series.


When Bookworm came out, I was running Bullseye with this driver and 
never really had any graphical issues that I can remember. Upon 
upgrading to Bullseye, I eventually found out this driver was no 
longer available and I was stuck with nouveau. I was recommended the 
tesla 470 driver during the upgrade, but I never got it to work.


According to Nvidia's official website GT 710 is supported up to 470 
version. [1]
I've never tried to install "tesla" flavor of nvidia-driver, so I can't 
suggest anything about it.
When I needed a specific version I always build a simple backport [2] 
from "testing" repo, so I'd try to do the same for a 470.199.02 version, 
which is now available from "oldstable-proposed-updates" repo.
You can also try the official package from Nvidia website, at least to 
test if 470 version works and solves your issue with freezing.


The nouveau driver would be fine, since I don't do anything 
graphics-intensive like gaming and I get the bonus of being able to 
run Wayland, which is cool. Except that I started noticing constant 
freezes for no apparent reason. They usually take a few hours to 
happen, sometimes I can spend a day or maybe two without freezes, but 
they _keep happening_. I'm gonna spare you of the details of the 
effort I spent trying to investigate this problem, but what I found 
out is:


  * It happens in any of the kernels I have installed: 6.1, 6.0, 5.11, 4.9
  * It happens in Plasma, and it doesn't matter if I'm running X11 or
Wayland
  * It _seems_ not to happen under GNOME Wayland, but I think it
happens under GNOME X11
  * It appears to be related with screen sharing, because:
  o When I was using Plasma, I had the impression that the
graphics froze way more often when I was sharing my screen
  o Screen sharing doesn't work on GNOME Wayland (maybe related to
why it didn't freeze yet)
  o The one time where I tried to use GNOME X11, it froze
immediately after I tried to share my screen
  * The kernel logs seem to indicate some missing firmware right
before the graphics card freezes (and indeed some firmware is missing)

I had issues very similar to your description. Even made a bug report 
[3] about it.
I don't have any instability or "freezing" issues with recent kernel and 
driver versions, but I've also disabled hardware acceleration for my 
browser as a workaround and haven't actually checked if the issue is gone.
Somehow it was all tied to video hardware acceleration and DE 
Compositor, which I kept disabled. I was able to repro the issue quite 
reliably, back then.
I've thought this workaround was better than downgrading driver version 
to 416 and stick to it forever.
Now I've upgraded system to Bookworm, and use 525.105.17 version, 
delaying update to an up-to-date version from stable, and haven't seen 
"Xid" errors or "freezes" for many months, Compositor is enabled, games 
and movies play without any issues.


Screen sharing is essential for me to work, so this is really not a 
great situation to be in. For comparison, my laptop, which was also 
updated to Bookworm in the same day I updated this PC, had none of the 
problems I'm describing here, since it has an integrated intel 
graphics card.


I appreciate that Nvidia cards can be very problematic and that the 
one card I own is really old, but I didn't have this problem until 
updating and I know this card can work just fine with the right 
driver. Unfortunately, right now, nouveau is not adequate for me to 
work, but the driver that does work is not officially available. What 
should I do?


  * Should I try updating to a newer kernel to see if nouveau got fixed?

I haven't checked, but I think nouveau didn't had a commit in years. It 
is an abandon-ware for me.



  * Should I try installing the missing firmware?


I'd try that, at least to see if it solves the issue.
Keep in mind, non-free firmware was separated to another repo in 
Bookworm. [4]



[1] https://www.nvidia.com/Download/driverResults.aspx/205995/en-us/
[2] https://wiki.debian.org/SimpleBackportCreation
[3] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1016542
[4] https://wiki.debian.org/SourcesList

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Bookworm boot stacks with black screen after NVIDIA driver installed.

[Alexander V. Makartsev]

On 16.08.2023 04:06, Juan R.D. Silva wrote:

Hi folks.

Fresh Bookworm install on Dell M4800 Precision with i7-4810MQ CPU @ 
2.80GHz and NVIDIA Quadro K2100M graphic card. No problems install, 
the system works with the default nouveau driver well enough.


My video card is supported according to NVIDIA. 
Information on official Nvidia website suggests otherwise. Last driver 
version that supports your VGA is 418 release. [1]

Newer driver versions don't mention K2100M on "Supported Products" list.

Debian nvidia-detect advised installation of nvidia-tesla-470-driver 
available in repo. After installing it the system boot stacks 
somewhere in the middle with a weird black screen. No access to TTYs, 
no cursor, no reaction to keyboard. I tried to boot with "nomodeset" 
in GRUB with no result.
Installation of drivers newer than 418 version wouldn't work. The only 
thing they will do is blacklist "nouveau".

Version 418 is available in Buster which is current "oldoldstable". [2]
You can try to backport driver from "oldoldstable" to Bookworm. It might 
be an impossible task due to possible incompatibilities with kernel 6.x

There is also an option to install official driver package from Nvidia.
Or you can install and use Buster.


[1] https://www.nvidia.com/download/driverResults.aspx/153717/en-us/
[2] https://wiki.debian.org/DebianOldOldStable

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Oddity when accessing Courier IMAP from Android

[Alexander V. Makartsev]

On 11.08.2023 19:09, Roberto C. Sánchez wrote:

I am running a Courier IMAP server on Bookworm.

When accessing from mutt, Thunderbird, and Android (GMail app),
everything works fine. This is over SSL, with a server certificate I
have issued from my own CA and with the CA root cert distributed to all
client devices which access the IMAP server.

However, on one particular Android device, I have started to observe
that my account no longer refreshes and when I look in the server logs,
I see a LOGIN, followed by "Unexpected SSL connection shutdown" followed
immediately by a DISCONNECT.

I have seen various postings online about this, but they are mostly from
10+ years ago and have to do with things like people forgetting to trust
the certificate and things of that nature.

I'm wondering if anyone has any sort of thought on what might result in
a single device experiencing this problem when other devices are able to
access the server and the account without issue.

Regards,

-Roberto

It could go wrong for many reasons. Could be a broken SSL\TLS 
implementation on that Android device, which usually fixed by firmware 
updates.
There also support for different ciphers and SSL\TLS versions. Many of 
them are outdated by today's standards and your Android device is 
essentially frozen in time.
I have functional Android device which is no longer supported by Google 
services and refuses to connect to any websites via HTTPS, mostly 
because of outdated Root CA certificates, but it also refuses to connect 
to my Samba server for some reason.
I suggest to look for a way to enable debug mode for Courier server and 
look for clues in logs.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: why bookworm isn't called deb12?

[Alexander V. Makartsev]

On 07.07.2023 01:23, hlyg wrote:

it seems natural to me to use deb12 for debian 12

deb for debian as in file name extension of package

it follows Windows naming style: win7, win8 ...

but others don't think so, i google with deb12, few means debian

in past 20 years few call it debN (N=1,2,3...)

why few are interested in saving 4 characters (ian )?


I'd save Ian if I could.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: cannot install odcb mariadb in bookworm

[Alexander V. Makartsev]

On 06.07.2023 03:09, John Covici wrote:

On Wed, 05 Jul 2023 15:47:39 -0400,
Alexander V. Makartsev wrote:

...
It's also a mystery why OP is trying to install "everything
odbc", I doubt they need a driver for MSSQL along side with
drivers for MariaDB and Postgres.
So I'm on a path to figure out what could be wrong with the OP's
system and what is their final goal.

I am upgrading freepbx...

You've probably used this tutorial [1] to install FreePBX?
If so, I think you can try to remove "/etc/odbcinst.ini" file and retry 
reinstallation of packages, because this file should be populated by 
odbcinst script automatically and maybe something in it prevents script 
from working.
If that still doesn't help you've probably better asking for help on 
FreePBX forums. [2]



[1] 
https://wiki.freepbx.org/display/FDT/How+to+Install+FreePBX+16+on+Debian+11+with+Asterisk+16

[2] https://community.freepbx.org/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: cannot install odcb mariadb in bookworm

[Alexander V. Makartsev]

On 06.07.2023 00:18, Greg Wooledge wrote:

On Thu, Jul 06, 2023 at 12:07:27AM +0500, Alexander V. Makartsev wrote:

I'm asking because package names you trying to install have prefixes like
"0-...", "1-...", etc.
Are you following some tutorial, or you've manually downloaded ".deb" files
from somewhere?

I've seen that too, on my systems.  It's something that apt-get is
doing automatically, and it does not reflect the actual filenames of
the .deb files as they sit in /var/cache/apt/archives/.

Perhaps apt-get creates a symlink farm temp directory somewhere, with
these filenames as symlinks pointing to the real .deb files, for some
internal purpose.  I don't know.  It never felt important enough for me
to try to figure it out.

It sure makes more difficult to compare the outputs...


I doubt it has any relevance to the OP's problem.  I'd focus on this
error instead:

Setting up tdsodbc:amd64 (1.3.17+ds-2) ...
odbcinst: SQLInstallDriverEx failed with Unable to find component
name.
dpkg: error processing package tdsodbc:amd64 (--configure):
  installed tdsodbc:amd64 package post-installation script subprocess
  returned error exit status 1
Even though error message seems obvious, "tdsodbc" package and all other 
odbc related packages install on my system without errors and I still 
can't reproduce the OP's problem.


   ...
   Setting up tdsodbc:amd64 (1.3.17+ds-2) ...
   odbcinst: Driver installed. Usage count increased to 1.
    Target directory is /etc
   ...

It's also a mystery why OP is trying to install "everything odbc", I 
doubt they need a driver for MSSQL along side with drivers for MariaDB 
and Postgres.
So I'm on a path to figure out what could be wrong with the OP's system 
and what is their final goal.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: cannot install odcb mariadb in bookworm

[Alexander V. Makartsev]

On 05.07.2023 23:23, John Covici wrote:

On Sat, 01 Jul 2023 06:16:33 -0400,
Alexander V. Makartsev wrote:

[1  ]
On 30.06.2023 03:11, John Covici wrote:

Hi.  I am trying to install odcb-mariadb in bookworm.  It was fine in
bullseye, but in bookworm I get the following error:
Unpacking odbc-mariadb (3.1.15-3) over (3.1.15-3) ...
Setting up odbc-mariadb (3.1.15-3) ...
odbcinst: SQLInstallDriverEx failed with Unable to find component
name.
dpkg: error processing package odbc-mariadb (--configure):

How to fix?

Thanks in advance for any suggestions.


No errors on my system.
You might need to add "bookworm-proposed-updates" repo to
"/etc/apt/sources.list" and
reinstall "odbcinst" "unixodbc-common" "odbc-mariadb" packages like so:

$ sudo apt update
$ sudo apt install --reinstall unixodbc-common odbcinst odbc-mariadb

Backup "/etc/odbc.ini" and "/etc/odbcinst.ini" files before
re-installation just in case.


Diid not seem to work -- I got the following:

Reading changelogs... Done
Preconfiguring packages ...
(Reading database ... 140553 files and directories currently
installed.)
Preparing to unpack .../0-unixodbc-common_2.3.11-2+deb12u1_all.deb ...
Unpacking unixodbc-common (2.3.11-2+deb12u1) over (2.3.11-2) ...
Preparing to unpack .../1-libodbcinst2_2.3.11-2+deb12u1_amd64.deb ...
Unpacking libodbcinst2:amd64 (2.3.11-2+deb12u1) over (2.3.11-2) ...
Preparing to unpack .../2-odbcinst1debian2_2.3.11-2+deb12u1_amd64.deb
...
Unpacking odbcinst1debian2:amd64 (2.3.11-2+deb12u1) over (2.3.11-2)
...
Preparing to unpack .../3-odbcinst_2.3.11-2+deb12u1_amd64.deb ...
Unpacking odbcinst (2.3.11-2+deb12u1) over (2.3.11-2) ...
Preparing to unpack .../4-libodbc2_2.3.11-2+deb12u1_amd64.deb ...
Unpacking libodbc2:amd64 (2.3.11-2+deb12u1) over (2.3.11-2) ...
Preparing to unpack .../5-libodbc1_2.3.11-2+deb12u1_amd64.deb ...
Unpacking libodbc1:amd64 (2.3.11-2+deb12u1) over (2.3.11-2) ...
Preparing to unpack .../6-libodbccr2_2.3.11-2+deb12u1_amd64.deb ...
...

Any ideas?


Are you using any 3rd-party apt repositories?
I'm asking because package names you trying to install have prefixes 
like "0-...", "1-...", etc.
Are you following some tutorial, or you've manually downloaded ".deb" 
files from somewhere?


Please include the exact command you've trying to run and its entire 
output, not just snippets of it, in the next reply.

Additionally, get the output from this command:

   $ grep -rE "^deb " /etc/apt/sources.list*

You can use paste [1] service and put just the link in reply, if the 
output is too long.



[1] https://paste.debian.net/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Raid Array and Changing Motherboard

[Alexander V. Makartsev]

On 02.07.2023 22:23, Mick Ab wrote:


I have a software RAID 1 array of two hard drives. Each of the two 
disks contains the Debian operating system and user data.


I am thinking of changing the motherboard because of problems that 
might be connected to the current motherboard. The new motherboard 
would be the same make and model as the current motherboard.


Would I need to recreate the RAID 1 array for the new motherboard I.e. 
re-initialise the current RAID 1 disks and repopulate the disks with 
data or can I just set up the software RAID on the new motherboard 
without affecting the current data on the RAID 1 drives ?


It's hard to tell what exactly will happen, because it depends on 
BIOS/Firmware of the motherboard, even though there is a special 
metadata record on each disk, which contains role of the disk and 
configuration of the RAID array. I predict two outcomes:
1. Two disks connected to a new motherboard will be recognized by 
BIOS/Firmware right away after you switch controller mode from AHCI to 
RAID, and appear as existing RAID1 array.
2. Two disks connected to a new motherboard will appear as two normal 
disks and won't be recognized as a RAID1 array, asking you to 
create\init array.


In case #2 data on disks will be lost, so before you do any 
manipulations make and verify backups.
Usually BIOS RAID software is very basic and won't allow to preserve 
current data on disks, or select a role (primary/secondary) for the 
disks, or create incomplete RAID1 array using only one disk to allow to 
copy data over from the second disk.


If you happen to have any other two old disks on hand, I suggest you to 
experiment with those on current motherboard, i.e. create an additional 
new RAID1 array and see if that array stays intact after simulated disks 
"transfer".
You can simulate disks transfer by powering of the computer, 
disconnecting the test disks and check if test RAID1 array still listed.
If test array will be listed and report two test disks missing then 
array information is also recorded in BIOS and this array information 
won't be on a new motherboard.
However, if there won't be any information about test array, then it 
should appear when you reconnect test disks and data on test disks 
should be intact.


There could be also a manual available from motherboard's manufacturer 
which could give some clues about what is possible and what would happen.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: cannot install odcb mariadb in bookworm

[Alexander V. Makartsev]

On 01.07.2023 18:44, John Covici wrote:

OK, thanks much --what do I add to my sources list for the proposed
updates?  Do I need all the lines ending with main free etc. or just
one line?

It is up to you. If you need to solve the "odbc mariadb" problem, I 
think selecting just "main" will be enough, eg.:

deb https://deb.debian.org/debian/ bookworm-proposed-updates main
deb-src https://deb.debian.org/debian/ bookworm-proposed-updates main

Or you can add all other sections like so:
deb https://deb.debian.org/debian/ bookworm-proposed-updates main 
contrib non-free non-free-firmware
deb-src https://deb.debian.org/debian/ bookworm-proposed-updates main 
contrib non-free non-free-firmware


Consult Debian wiki for additional info:
https://wiki.debian.org/SourcesList

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: cannot install odcb mariadb in bookworm

[Alexander V. Makartsev]

On 30.06.2023 03:11, John Covici wrote:

Hi.  I am trying to install odcb-mariadb in bookworm.  It was fine in
bullseye, but in bookworm I get the following error:
Unpacking odbc-mariadb (3.1.15-3) over (3.1.15-3) ...
Setting up odbc-mariadb (3.1.15-3) ...
odbcinst: SQLInstallDriverEx failed with Unable to find component
name.
dpkg: error processing package odbc-mariadb (--configure):

How to fix?

Thanks in advance for any suggestions.


No errors on my system.
You might need to add "bookworm-proposed-updates" repo to 
"/etc/apt/sources.list" and

reinstall "odbcinst" "unixodbc-common" "odbc-mariadb" packages like so:

   $ sudo apt update
   $ sudo apt install --reinstall unixodbc-common odbcinst odbc-mariadb

Backup "/etc/odbc.ini" and "/etc/odbcinst.ini" files before 
re-installation just in case.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

debchange still wants to build for bullseye-backports after upgrade to Bookworm

[Alexander V. Makartsev]

Hello.

I've successfully upgraded to Bookworm recently and trying to build a 
backport package.
But the usual "$ debchange --bpo" still wants to build for 
"bullseye-backports" and modifies "debian/changelog" by adding 
"~bpo11+1" and "bullseye-backports;"

instead of expected "~bpo12+1" and "bookworm-backports;"
Is there anything to check on my system or this is a bug in "dch" and/or 
"devscripts" package?


   $ dch --help | grep -iEA2 " --bpo"
  --bpo
 Increment the Debian release number for a backports upload
 to "bullseye-backports"

   $ uname -a
   Linux hostname 6.1.0-9-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.27-1
   (2023-05-08) x86_64 GNU/Linux

   $ cat /etc/os-release | grep -iE "_ID|_CODE"
   VERSION_ID="12"
   VERSION_CODENAME=bookworm

   $ apt-show-versions devscripts
   devscripts:all not installed
   devscripts:amd64/bookworm 2.23.4 uptodate
   devscripts:i386 not installed



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: unzip files bigger than 4 GB

[Alexander V. Makartsev]

On 14.06.2023 23:19, Van Snyder wrote:
unzip v 6.0 (the version delivered with Debian 10) doesn't work with 
files bigger than 2^32 bytes.


Is there an alternative program to do it?


"7zip" is the best. It supports multiple formats and cross-platform.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: NetworkManager, Iphone, Bullseye, and Bookworm

[Alexander V. Makartsev]

On 27.05.2023 22:43, Charles Curley wrote:

I recently upgraded from an iPhone 8 to a 14. All of my various WiFi
capable computers worked great with the iPhone 8's Personal Hotspot.
However, the new iPhone does not appear to like Bullseye. Two computers
show the same symptoms: the iPhone does not seem to like the password
previously stored and known to work.

On one, I even moved the old connection file aside and tried creating a
new connection file. I entered the password, and had repeated requests
to provide the password.

However, I have Bookworm on an ancient IBM R51. That got on the
Personal Hotspot immediately with no fuss. I also ran a recent live CD
of Bookworm on one of the two failing Bullseye machines, and that worked
immediately.

On one of the Bullseye machines, I upgraded NetworkManager from
backports. No go. So, is there something in that mess of dependencies
that NetworkManager rides herd on that I should upgrade, preferably from
Bullseye backports? Other thoughts? (Aside from install Bookworm
immediately.)


Probably, an updated "wpasupplicant" package is what you need.
The version in "bullseye-backports" is slightly older, so you might need 
to build a backport from sources in "bookworm".


   $ rmadison wpasupplicant
   wpasupplicant | 2:2.7+git20190128+0c1e29f-4~bpo9+2  |
   stretch-backports  | amd64, arm64, armel, armhf, i386, mips,
   mips64el, mipsel, ppc64el, s390x
   wpasupplicant | 2:2.7+git20190128+0c1e29f-6+deb10u3 |
   oldstable  | amd64, arm64, armel, armhf, i386, mips,
   mips64el, mipsel, ppc64el, s390x
   wpasupplicant | 2:2.9.0-21  |
   stable | amd64, arm64, armel, armhf, i386, mips64el,
   mipsel, ppc64el, s390x
   wpasupplicant | 2:2.10-8~bpo11+2    |
   bullseye-backports | amd64, arm64, armel, armhf, i386, mips64el,
   mipsel, ppc64el, s390x
   wpasupplicant | 2:2.10-12   |
   testing    | amd64, arm64, armel, armhf, i386, mips64el,
   mipsel, ppc64el, s390x
   wpasupplicant | 2:2.10-12   |
   unstable   | amd64, arm64, armel, armhf, i386, mips64el,
   mipsel, ppc64el, s390x


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to download source package using only console?

[Alexander V. Makartsev]

On 15.05.2023 15:19, Vincent Lefevre wrote:

On 2023-05-15 10:25:45 +0500, Alexander V. Makartsev wrote:


I see. That explains why I can request source package
"golang-github-xenolf-lego/testing" directly and get the right one.
So, in my case, I won't be able to reliably get a source package(-s) from
"testing" if I don't add "deb" part of "testing" to "sources.list", which
could be a different can of worms...
Is this something for me to just be aware of and leave it as is now, or is
there more elegant solution?

Well, if you don't want the "deb" part, you need to provide the
name of the source package directly to "apt source". If you do that
frequently, you can write a script. There are 2 solutions:
* A remote request, e.g. with rmadison.
* Something based of "apt cache show". From that, you can get the
   source package, then call "apt source" with the source package.
   But note that this is only a heuristic; if the name of the source
   package has changed from stable to testing, this won't work.

Note that adding the "deb" part shouldn't be much an issue (except
noise, e.g. with "apt cache show", which would give output for both
stable and testing). If you want to make sure that a package from
testing won't be installed by mistake (by apt), I suppose that you
can use apt preferences with a negative priority for testing to
prevent such an installation:

Package: *
Pin: release a=testing
Pin-Priority: -1

(not tried). See the apt_preferences(5) man page for details.

It looks like adding the "deb" part with apt pinning is the best option.
Even with pinning in effect, packages from "testing" still could be 
installed, but only if they were manually requested.

Now the command "$ apt source lego/testing" is working properly.
Thanks for suggestions, Vincent.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to download source package using only console?

[Alexander V. Makartsev]

On 15.05.2023 05:43, Vincent Lefevre wrote:

On 2023-05-14 14:17:05 +0500, Alexander V. Makartsev wrote:

[...]
I think you haven't noticed that I requested for "4.9.1-1" version from
"testing" specifically,

You can't. You can either request some given version, e.g. 4.9.1-1
(but this will work only if it can be found from your local database),
or the version from some given distribution, e.g. "testing".

But my point is that your database is obsolete, because if you ask
the version from testing, apt thinks that it is 3.2.0-3.1, while it
should be 4.9.1-1. You need to fix that.
What is the best approach to fix that? Keep in mind, I only need a 
source package(-s) from "testing".

So, just to be safe, "deb" source for "testing" was commented out:

   $ cat /etc/apt/sources.list | grep -iE "testing"
   #deb https://deb.debian.org/debian/ testing main contrib non-free
   deb-src https://deb.debian.org/debian/ testing main contrib non-free



hence why the command was "$ apt source lego/testing" not just "$ apt source
lego".
There is no reason for building a backport package for "stable" using a
source package from "stable"...

I've changed all my repo mirrors to "deb.debian.org" suspecting the previous
mirror I used was somehow out-of-date. Why is my output differ from yours?

$ apt-show-versions -a lego
lego:amd64 3.2.0-3.1+b5 bullseye deb.debian.org
No proposed-updates version
No stable-updates version
No testing version
No unstable version
lego:amd64 not installed
lego:i386 3.2.0-3.1+b5 bullseye deb.debian.org
No proposed-updates version
No stable-updates version
No testing version
No unstable version
lego:i386 not installed

After changing your sources.list, you need an "apt update" again.
Of course, I always do "$ sudo apt update" after changing apt config 
files and before any package manipulation.



Yet "rmadison" reports there is a version "4.9.1-1" available in "testing":

$ rmadison lego
lego   | 0.3.1-5+b13   | oldstable  | amd64, arm64, armel,
armhf, i386, mips, mips64el, mipsel, ppc64el, s390x
lego   | 3.2.0-3.1+b5  | stable | amd64, arm64, armel,
armhf, i386, mips64el, mipsel, ppc64el, s390x
lego   | 4.9.1-1   | testing    | amd64, arm64, armel,
armhf, i386, mips64el, mipsel, ppc64el, s390x
lego   | 4.9.1-1   | unstable   | amd64, arm64, armel,
armhf, i386, mips64el, mipsel, ppc64el, s390x

I suspect "apt-show-versions" output is inconsistent because I only request
"deb-src" from "testing" in "sources.list", as I've shown before.

Probably. "apt-show-versions" considers only the binary packages
(but "lego" is only a binary package).
I see. "rmadison" utility is checking out repos directly, where as 
"apt-show-versions" rely on local database information.



[...]

I did some additional research and I think I got it.
"lego" package is special because its source package is named differently:

Yes, as said by apt above:

   Picking 'golang-github-xenolf-lego' as source package instead of 'lego'

[...]

But why "apt" doesn't play along, since it knows the source package for
"lego" has different name, but ignores the "testing" part of the request?

$ apt source lego/testing
Reading package lists... Done
Picking 'golang-github-xenolf-lego' as source package instead of 'lego'
E: Can not find version '3.2.0-3.1' of package 'lego'
E: Unable to find a source package for golang-github-xenolf-lego

Looks like an "apt" bug to me.

Probably not. You are doing a request on a binary package (since
"lego" is not a source package). The translation from the binary
package to the source package depends on the particular version of
the binary package. So lego/testing will correspond to the binary
package from testing, which is 3.2.0-3.1+b5 in your case (because
of your obsolete database due to the missing "deb" for testing in
sources.list). Then apt translates this to the source package (of
the same version) golang-github-xenolf-lego 3.2.0-3.1, which is
unknown on your machine because the deb-src database is up-to-date
(contrary to the deb database). Something like that.

I see. That explains why I can request source package 
"golang-github-xenolf-lego/testing" directly and get the right one.
So, in my case, I won't be able to reliably get a source package(-s) 
from "testing" if I don't add "deb" part of "testing" to "sources.list", 
which could be a different can of worms...
Is this something for me to just be aware of and leave it as is now, or 
is there more elegant solution?


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to download source package using only console?

[Alexander V. Makartsev]

On 14.05.2023 10:06, Vincent Lefevre wrote:

On 2023-05-14 00:15:39 +0500, Alexander V. Makartsev wrote:

Hello, fellow Debian users.

When I need to build a backport of a package, I sometimes find it difficult
to obtain actual source package(-s) from Debian repos using console.
Following advice from a wiki page [1], after "apt update", doesn't do it:

$ apt source lego/testing
Reading package lists... Done
Picking 'golang-github-xenolf-lego' as source package instead of 'lego'
E: Can not find version '3.2.0-3.1' of package 'lego'
E: Unable to find a source package for golang-github-xenolf-lego

zira:~> apt-show-versions -a lego
lego:amd64 3.2.0-3.1+b5 stableftp.debian.org
No stable-updates version
lego:amd64 4.9.1-1  testingftp.debian.org
lego:amd64 4.9.1-1  unstableftp.debian.org
No experimental version
lego:amd64 not installed

Indeed, 3.2.0-3.1 is no longer the testing version. Your database
seems to be out-of-date.

I think you haven't noticed that I requested for "4.9.1-1" version from 
"testing" specifically,
hence why the command was "$ apt source lego/testing" not just "$ apt 
source lego".
There is no reason for building a backport package for "stable" using a 
source package from "stable"...


I've changed all my repo mirrors to "deb.debian.org" suspecting the 
previous mirror I used was somehow out-of-date. Why is my output differ 
from yours?


   $ apt-show-versions -a lego
   lego:amd64 3.2.0-3.1+b5 bullseye deb.debian.org
   No proposed-updates version
   No stable-updates version
   No testing version
   No unstable version
   lego:amd64 not installed
   lego:i386 3.2.0-3.1+b5 bullseye deb.debian.org
   No proposed-updates version
   No stable-updates version
   No testing version
   No unstable version
   lego:i386 not installed

Yet "rmadison" reports there is a version "4.9.1-1" available in "testing":

   $ rmadison lego
   lego   | 0.3.1-5+b13   | oldstable  | amd64, arm64, armel,
   armhf, i386, mips, mips64el, mipsel, ppc64el, s390x
   lego   | 3.2.0-3.1+b5  | stable | amd64, arm64, armel,
   armhf, i386, mips64el, mipsel, ppc64el, s390x
   lego   | 4.9.1-1   | testing    | amd64, arm64, armel,
   armhf, i386, mips64el, mipsel, ppc64el, s390x
   lego   | 4.9.1-1   | unstable   | amd64, arm64, armel,
   armhf, i386, mips64el, mipsel, ppc64el, s390x

I suspect "apt-show-versions" output is inconsistent because I only 
request "deb-src" from "testing" in "sources.list", as I've shown before.


Here is another example package that works as expected:

   $ rmadison roundcube
   roundcube  | 1.3.17+dfsg.1-1~deb10u2 | oldstable    |
   source, all
   roundcube  | 1.4.13+dfsg.1-1~deb11u1~bpo10+1 | buster-backports |
   source, all
   roundcube  | 1.4.13+dfsg.1-1~deb11u1 | stable   |
   source, all
   roundcube  | 1.6.1+dfsg-1    | testing  |
   source, all
   roundcube  | 1.6.1+dfsg-1    | unstable |
   source, all

   $ apt-show-versions -a roundcube
   roundcube:all 1.4.13+dfsg.1-1~deb11u1 bullseye  deb.debian.org
   roundcube:all 1.4.13+dfsg.1-1~deb11u1 bullseye-security deb.debian.org
   No proposed-updates version
   No stable-updates version
   No testing version
   No unstable version
   roundcube:all not installed

   $ apt source roundcube/testing
   Reading package lists... Done
   Selected version '1.6.1+dfsg-1' (testing) for roundcube
   ...


I did some additional research and I think I got it.
"lego" package is special because its source package is named differently:

   $ rmadison golang-github-xenolf-lego
   golang-github-xenolf-lego | 0.3.1-5   | oldstable  | source
   golang-github-xenolf-lego | 3.2.0-3.1 | stable | source
   golang-github-xenolf-lego | 4.9.1-1   | testing    | source
   golang-github-xenolf-lego | 4.9.1-1   | unstable   | source
   golang-github-xenolf-lego | 4.9.1-1   | unstable-debug | source

   $ apt source golang-github-xenolf-lego/testing
   Reading package lists... Done
   Selected version '4.9.1-1' (testing) for golang-github-xenolf-lego
   ...

But why "apt" doesn't play along, since it knows the source package for 
"lego" has different name, but ignores the "testing" part of the request?


   $ apt source lego/testing
   Reading package lists... Done
   Picking 'golang-github-xenolf-lego' as source package instead of 'lego'
   E: Can not find version '3.2.0-3.1' of package 'lego'
   E: Unable to find a source package for golang-github-xenolf-lego


Looks like an "apt" bug to me.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

How to download source package using only console?

[Alexander V. Makartsev]

Hello, fellow Debian users.

When I need to build a backport of a package, I sometimes find it 
difficult to obtain actual source package(-s) from Debian repos using 
console.

Following advice from a wiki page [1], after "apt update", doesn't do it:

   $ apt source lego/testing
   Reading package lists... Done
   Picking 'golang-github-xenolf-lego' as source package instead of 'lego'
   E: Can not find version '3.2.0-3.1' of package 'lego'
   E: Unable to find a source package for golang-github-xenolf-lego

Any other seemingly intended ways also fail in a similar fashion:

   $ dget lego=4.9.1-1
   dget: no hostnames in apt-cache policy lego for version 4.9.1-1 found

   $ apt-get source lego -t testing
   Reading package lists... Done
   E: The value 'testing' is invalid for APT::Default-Release as such a
   release is not available in the sources
   E: Unable to find a source package for

Trying to do the same for another package seems to work:

   $ apt source ipcalc/testing
   Reading package lists... Done
   Selected version '0.42-2' (testing) for ipcalc
   Need to get 33,7 kB of source archives.
   Get:1 https://mirror.yandex.ru/debian testing/main ipcalc 0.42-2
   (dsc) [1 692 B]
   Get:2 https://mirror.yandex.ru/debian testing/main ipcalc 0.42-2
   (tar) [25,9 kB]
   Get:3 https://mirror.yandex.ru/debian testing/main ipcalc 0.42-2
   (diff) [6 144 B]
   Fetched 33,7 kB in 1s (52,5 kB/s)
   dpkg-source: info: extracting ipcalc in ipcalc-0.42
   dpkg-source: info: unpacking ipcalc_0.42.orig.tar.gz
   dpkg-source: info: unpacking ipcalc_0.42-2.debian.tar.xz
   dpkg-source: info: using patch list from debian/patches/series
   dpkg-source: info: applying 01-paths.patch


So why those fail for a "lego" package and is there a way to solve this 
once and for all?
I know I can go to a packages website [2] and manually download ".dsc" 
file and feed it to "dget" utility, or download source files directly 
from said website, but there has to be a better way.


Some useful info:

   $ cat /etc/apt/sources.list | grep -iE "testing"
   #deb https://mirror.yandex.ru/debian/ testing main contrib non-free
   deb-src https://mirror.yandex.ru/debian/ testing main contrib non-free

   $ rmadison lego
   lego   | 0.3.1-5+b13   | oldstable  | amd64, arm64, armel,
   armhf, i386, mips, mips64el, mipsel, ppc64el, s390x
   lego   | 3.2.0-3.1+b5  | stable | amd64, arm64, armel,
   armhf, i386, mips64el, mipsel, ppc64el, s390x
   lego   | 4.9.1-1   | testing    | amd64, arm64, armel,
   armhf, i386, mips64el, mipsel, ppc64el, s390x
   lego   | 4.9.1-1   | unstable   | amd64, arm64, armel,
   armhf, i386, mips64el, mipsel, ppc64el, s390x

   $ uname -a
   Linux host0 5.10.0-22-amd64 #1 SMP Debian 5.10.178-3 (2023-04-22)
   x86_64 GNU/Linux



[1] https://wiki.debian.org/SimpleBackportCreation
[2] https://packages.debian.org/bookworm/lego

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: What do all those "* * *" mean on a traceroute log?

[Alexander V. Makartsev]

On 13.04.2023 00:00, Albretch Mueller wrote:

  Yes, but should it happen on every hop? In my case it happens while I
am trying to reach every site and from wherever I have the chance to
get some relatively decent Internet access?
There is a chance your trace packets were filtered (rate-limited), 
because by default "traceroute" sends them without delay and remote 
hosts could "see" them as flood.
Try to test same route again, but with a send delay set to a reasonable 
1 second using "-z" parameter, like so:

    # traceroute -z 1 8.8.8.8



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Pinning not working?!

[Alexander V. Makartsev]

On 04.04.2023 00:12, Thomas Schweikle wrote:


does not seen to work at all, since the 4.1-2 package has priority 500 
but if pinning would work it should have 1000. What is wrong here?

It works for me.
    Without pinning:
$ apt-cache policy nvidia-driver
nvidia-driver:
  Installed: 470.161.03-1
  Candidate: 470.161.03-1
  Version table:
 *** 470.161.03-1 500
    500 https://mirror.yandex.ru/debian bullseye/non-free amd64 
Packages

    100 /var/lib/dpkg/status
 470.103.01-1~bpo11+1 100
    100 https://mirror.yandex.ru/debian bullseye-backports/non-free 
amd64 Packages


    Created .pref file:
$ cat /etc/apt/preferences.d/nvidia-driver.pref
Package: /nvidia-driver/
Pin: version /470.103/
Pin-Priority: 1000

    With pinning after "apt update":
$ apt-cache policy nvidia-driver
nvidia-driver:
  Installed: 470.161.03-1
  Candidate: 470.103.01-1~bpo11+1
  Version table:
 *** 470.161.03-1 500
    500 https://mirror.yandex.ru/debian bullseye/non-free amd64 
Packages

    100 /var/lib/dpkg/status
 470.103.01-1~bpo11+1 1000
    100 https://mirror.yandex.ru/debian bullseye-backports/non-free 
amd64 Packages


Try using proper regexp syntax, to pin candidates.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: bendel.debian.org untrusted certificate

[Alexander V. Makartsev]

On 12.03.2023 09:55, jeremy ardley wrote:


Back on topic I have solved getting letsencrypt certificates accepted 
by postfix , but not certificates generated by Debian SMTP CA


Received: from edge.bronzemail.com 
(2403-5800-c000-1b7-f3d4-d970-ca28-bf4f.ip6.aussiebb.net 
[IPv6:2403:5800:c000:1b7:f3d4:d970:ca28:bf4f])

(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)
 key-exchange X25519 server-signature RSA-PSS (2048 bits) 
server-digest SHA256

 client-signature RSA-PSS (2048 bits) client-digest SHA256)
(Client CN "edge.bronzemail.com", Issuer "R3" (verified OK))
by mail.bronzemail.com (Postfix) with ESMTPS id B38C6860225
for ; Sun, 12 Mar 2023 12:50:55 +0800 (AWST)
Authentication-Results: mail.bronzemail.com;
dkim=pass (2048-bit key; unprotected) header.d=ardley.org 
header.i=@ardley.org header.a=rsa-sha256 header.s=default 
header.b=OfZ0VNf6;

dkim-atps=neutral
Received: from bendel.debian.org (bendel.debian.org [82.195.75.100])
(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)
 key-exchange X25519 server-signature RSA-PSS (2048 bits) 
server-digest SHA256

 client-signature RSA-PSS (2048 bits) client-digest SHA256)
(Client CN "clientcerts/bendel.debian.org", Issuer "Debian SMTP 
CA" (not verified))

by edge.bronzemail.com (Postfix) with ESMTPS id 691864037F
for ; Sun, 12 Mar 2023 12:50:53 +0800 (AWST)

It says "Issuer "Debian SMTP CA" (not verified)" because your client 
(mail server) doesn't have any means to verify certificate of this 
Certificate Authority.
"Debian SMTP CA" is self-hosted Certificate Authority. That means it is 
first of the chain, so it doesn't have any globally trusted CAs above it 
that could verify its legitimacy.
To make your client to trust it, you have to install its public 
certificate into trusted CA storage ( to "/etc/ssl/certs/" or where your 
client was configured to look for trusted CA certs ).


You can obtain CA certificate with "openssl" utility:
$ openssl s_client -tls1_3 -showcerts -starttls smtp -connect 
bendel.debian.org:25


First certificate, identified by "-BEGIN CERTIFICATE-" line, is 
always server's certificate, last is CA certificate. You can save it to 
external file, so its contents would look like this:

$ cat bendel.debian.org.issuer.crt
-BEGIN CERTIFICATE-
MIIE4jCCA8qgAwIBAgIJAKMp9hZf6dHhMA0GCSqGSIb3DQEBCwUAMIGmMQswCQYD
...
ogSbTuTG
-END CERTIFICATE-

You can additionally check this certificate for consistency:
$ openssl x509 -text -in ./bendel.debian.org.issuer.crt | grep -iE 
"issuer:|subject:|ca:true"


Now you can verify server's certificate using this CA certificate:
With it:
$ openssl s_client -tls1_3 -CAfile bendel.debian.org.issuer.crt 
-starttls smtp -connect bendel.debian.org:25 2>&1 | grep "Verification"

Verification: OK

Without it:
$ openssl s_client -tls1_3 -CAfile /etc/ssl/certs/ca-certificates.crt 
-starttls smtp -connect bendel.debian.org:25 2>&1 | grep "Verification"

Verification error: self signed certificate in certificate chain

Final step is to include "bendel.debian.org.issuer.crt" into 
configuration of your mail server, by installing it to "/etc/ssl/certs/" 
or by pointing directly to it.



Now I wonder why "bendel.debian.org" admins decided not to use free 
Let's Encrypt certificate and decided to use their own self-hosted CA..



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: does your Thunderbird for deb11 often become unresponsive?

[Alexander V. Makartsev]

On 03.03.2023 09:02, hlyg wrote:
at first i suspect Thunderbird(TB) try to access some sites that are 
blocked in china


i disable show Start Page when it launches

i disable check email automatically for each account

but it becomes unresponsive when i File->New->Message

i really don't know cause of unresponsiveness

PS: even if blocked sites cause trouble, it's bug of TB, it shall 
timeout after it can't access sites for some time

I don't have any issues with TB on a PC with operating system on SSD.
If I had to guess the culprit, it could be a failing HDD.
You should check "/var/log/syslog" for anything suspicious and\or 
"smartctl".
This also could be a symptom of resource starvation. Run "free -h" 
and\or "top" to see if there is any resource hogs.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: support for ASUS AC1200 USB-AC53 Nano wifi dongle

[Alexander V. Makartsev]

On 08.02.2023 20:55, Gary Dale wrote:

On 2023-02-08 09:07, Gary Dale wrote:


The journalctl command returns nothing.

That's strange. Is it possible you've forgot that pound ("#") sign means 
"run as root"?


Found a github repository that compiles on Bullseye at 
https://github.com/morrownr/88x2bu. Then it's a matter of doing the 
following as root


git clone https://github.com/morrownr/88x2bu
cd 88x2bu-20210702 ## date string may different
make clean
make
make install

then rebooting. The wifi dongle now shows in "ip addr".


Good to hear you've made it working.
You might want to walk an extra mile and setup DKMS [1][2] for it, so it 
will automatically re-compile and re-install itself after every kernel 
image update.



[1] https://manpages.debian.org/bullseye/dkms/dkms.8.en.html
[2] https://www.xmodulo.com/build-kernel-module-dkms-linux.html
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: support for ASUS AC1200 USB-AC53 Nano wifi dongle

[Alexander V. Makartsev]

On 08.02.2023 09:07, Gary Dale wrote:


I thought this would be easier than it's turned out to be. There are 
Internet posts going back years about support for this device but 
nothing recent - including a 5 year old Ubuntu post saying it works. 
Other wifi devices seem to be recognized out of the box or with a 
simple install of non-free firmware but not this one - at least not in 
Bullseye or Bookworm.


The adapter itself seems to be quite popular so I'm hoping someone can 
provide some clues on how to make it work


Thanks.

Your device should be based on "RTL8822B" chip from Realtek, so you need 
to install "firmware-realtek" package.
If after doing that you still didn't get a functioning network wifi 
adapter you might need to build driver kernel module. [1]
This is what I had to do to get USB Bluetooth adapter from Asus to work 
without issues, even though it is supported by kernel in "bullseye".


It is always the best to include extra information about your setup when 
you asking for help.

At least output from these commands would be a start:
    $ uname -a
    $ lsusb -v -t
    # journalctl -b 0 --no-pager | grep -iE "rtl|rtk_|firmware"

If the output is long you can use "paste" service [2] and send us a link.


[1] 
https://www.asus.com/ca-en/networking-iot-servers/adapters/all-series/usb-ac53-nano/helpdesk_download/?model2Name=USB-AC53-Nano

[2] https://paste.debian.net/
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: USB enumeration issue

[Alexander V. Makartsev]

On 23.01.2023 11:40, Matthew McAllister wrote:

Hi all,

Since I upgraded packages a couple weeks ago, whenever I start my PC, 
I have to wait 60 seconds for the kernel to enumerate USB devices. 
Here's the log:


[    8.815277] usb 1-5: device descriptor read/64, error -110
[   24.431295] usb 1-5: device descriptor read/64, error -110
[   24.943220] usb 1-5: new high-speed USB device number 3 using xhci_hcd
[   30.319491] usb 1-5: device descriptor read/64, error -110
[   45.935494] usb 1-5: device descriptor read/64, error -110
[   46.044772] usb usb1-port5: attempt power cycle
[   46.523221] usb 1-5: new high-speed USB device number 4 using xhci_hcd
[   51.323562] usb 1-5: Device not responding to setup address.
[   56.331406] usb 1-5: Device not responding to setup address.
[   56.539402] usb 1-5: device not accepting address 4, error -71
[   56.943221] usb 1-5: new high-speed USB device number 5 using xhci_hcd
[   61.743759] usb 1-5: Device not responding to setup address.
[   66.751609] usb 1-5: Device not responding to setup address.
[   66.959391] usb 1-5: device not accepting address 5, error -71
[   66.960945] usb usb1-port5: unable to enumerate USB device

This occurs when *no USB cables are plugged in*. The kernel is 
stalling the entire boot process to enumerate some internal USB hub, I 
assume.


My front USB-C is broken as far as I can tell, so I tried unplugging 
the header. The issue persisted.


The front USB 3.0 work correctly and I couldn't get the header 
unplugged anyways, so I didn't test if that was the issue.


Any ideas what might be going on? Kernel is 6.1.4-1.

Matthew

Start troubleshooting process by unplugging all USB devices, doesn't 
matter, if it's empty USB extension cable, external USB hub, or USB 
thumb drive.
If you did that already and don't have USB-anything plugged in and still 
have the same issue, then this is a hardware problem, not a software 
problem.
If you unplugged everything and can't reproduce the issue anymore, then 
you have faulty USB device, which should be tested one by one, to 
determine which one is the culprit.


There is one additional thing, if your "PC" is a laptop, then it is 
possible there is an internal device inside that uses USB bus to 
function, e.g. WiFi adapter, WWAN adapter, etc.

In that case you need to open laptop to remove these devices.
If your laptop is HP or Lenovo brand, you should look for publicly 
available Disassembly and Maintenance Manuals for your model on 
manufacturer's official website.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Cloning a disk: partclone?

[Alexander V. Makartsev]

On 19.01.2023 23:49, Tom Browder wrote:
On my main PC, I would like to clone my boot drive onto another disk 
for 2 reasons:


1. Use a larger disk for the main drive
2. Create an emergency recovery disk

A new Debian package to me is "partclone". Questions:

+ Can that be used for both purposes?
Yes. Keep in mind, that it will clone partitions (with filesystems) 
separately, so you have to create partition table on destination drive 
manually.
Procedure will depend on what partition scheme your source drive uses: 
MBR or GPT.
In case of MBR, you will have to install bootloader on a destination 
drive after cloning.
In case of GPT, cloning the EFI partition should be enough to boot from 
a destination drive.




+ Can it do a complete clone on an active disk? Or do I need a live CD 
or USB stick?

No, it can't. Partitions have to be unmounted for partclone to work.
Since you will be creating LiveCD on USB stick and going offline anyway 
it might be better (easier) to use Clonezilla for this task.
Clonezilla is customizable, has curses GUI, uses partclone under the 
hood and also automates other things, like partition creation on a 
destination drive, bootloader installation, etc.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Passwords

[Alexander V. Makartsev]

On 17.01.2023 13:51, DdB wrote:

Everyone (and their friend) seem to know, how to work around this, which
apparently is common debian knowledge (which is nice).

But somehow, i feel there could be more caring about avoiding to teach
future hackers by accident. Is this kind of lesson appropriate for a
users list? - I doubt it.
This is not hacking. How to reset password on your computer is the is 
most basic system administration knowledge everyone should know.
This information is described in system administration manual and mostly 
applies for any Linux-based system.


Since it is so easy to reset user password on a computer you have 
physical access to, and it doesn't matter if OS is Windows or Linux-based,
so making your user password so long and complex, to the point it could 
be forgotten, is so not necessary.
Even 4-6 digit password is enough to secure access to local user account 
and impossible to forget.


Long and complex passwords are only necessary for non-local user 
accounts, e.g. Web-sites, e-Mail accounts, or any accounts exposed to 
the Internet.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: why rfkill not list my wifi device

[Alexander V. Makartsev]

On 10.01.2023 12:33, lsg wrote:


On 1/10/23 15:26, Alexander V. Makartsev wrote:
For this adapter, according to Wireless Wiki [1], you should use 
universal driver "rtl8xxxu". See if it was loaded:

    $ sudo lsusb -t
...

Thank Alexander!

actually my usb wifi adapter works well in buster but rfkill doesn't

i want to get rfkill to work

In that case check if "rfkill" kernel module is actually loaded, because 
it is necessary for rfkill utility to function:

    $ lsmod | grep -iE "rfkill"

Use "modprobe" utulity to load it:
    $ sudo modprobe rfkill

It is possible that driver module (your current version) for this 
adapter simply doesn't support "rfkill" subsystem.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: why rfkill not list my wifi device

[Alexander V. Makartsev]

On 10.01.2023 09:54, lsg wrote:


Thank Charles! it's usb wifi adapter, lsusb shows

Bus 001 Device 005: ID 0bda:8179 Realtek Semiconductor Corp. 
RTL8188EUS 802.11n Wireless Network Adapter


rfkill shows nothing

For this adapter, according to Wireless Wiki [1], you should use 
universal driver "rtl8xxxu". See if it was loaded:

    $ sudo lsusb -t

Check out the logs for relevant messages:
    $ sudo grep -iE "rtl|firmware" /var/log/syslog

This should give you a hint about a firmware file needed for this adapter.
Probably it is a good idea in general to install firmware package for 
Realtek-based devices from "non-free" section of Debian repos:

    $ sudo apt install firmware-realtek


[1] https://wireless.wiki.kernel.org/en/users/drivers/rtl819x
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: off topic, has anyone here built marlin from src?

[Alexander V. Makartsev]

On 26.12.2022 18:15, gene heskett wrote:

Greetings all;
...
One path involves Visual Studio which does not seem to be available 
for debian, so it appears the platformio path is the one to follow. 
But step by step instructions are pretty slim.


Can anyone help get me started?

I don't have any experience with CNCs or MarlinFW, but from what I've 
read from documentation [1] [2] it should be pretty straight forward.
First things first you need to install VSCode (Visual Studio Code IDE) 
[3]. Which is not available in Debian repos so you have to trust 
Microsoft and add their external repos along with trusted gpg key.
After VSCode installation, you can proceed and install "PlatformIO" 
which is an extension plugin for VSCode. [4]
To do that, basically you need to press "Ctrl+Shift+X" inside VSCode and 
type "PlatformIO" into search bar inside left-side panel and click 
Install button.
After that you will have a ready-to-go toolkit for MarlinFW 
configuration and compilation, and proceed with further steps, as 
described in [4].



[1] https://marlinfw.org/docs/basics/install.html
[2] https://marlinfw.org/docs/configuration/configuration.html
[3] https://code.visualstudio.com/docs/setup/linux
[4] https://marlinfw.org/docs/basics/install_platformio_vscode.html
--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Self hosting solution for Christmas

[Alexander V. Makartsev]

On 24.12.2022 13:03, Andre Rodier wrote:

Hello everyone,

Here my present for Christmas: a new version of HomeBox, the self 
hosted email solution.


Feel free to drop comments, create issues, update the docs, etc.

I released this quickly before going on vacation, so you may find some 
issues. However, this is mostly stable, and the code is easy to modify.


Also, you can now add and remove components individually, with an 
Ansible playbook.


https://github.com/progmaticltd/homebox

...

Merry Christmas,
André

Impressive feature list. Does it use MySQL or PostgreSQL for 
settings\configuration\users storage?
I might give it a spin to take a peek under the hood and see how much of 
it could be customized and tailored for my needs.

And also to compare it to other projects like say iRedMail or MailCow.
Happy holidays!

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: How to investigate sudden shutdown?

[Alexander V. Makartsev]

On 14.12.2022 12:22, Tobias Diekershoff wrote:

Hey everyone,

perhaps someone of you can help me with tool / log file I have not investigated
so far. I have a Thinkpad with Debian Bullseye on it, running KDE/Plasma as main
desktop environment and from time to time it just turn off without prior
indication to do so.
Does it power off unexpectedly while you actively working on it, or you 
just see it powered off when you check it after a while?
These kinds of symptoms are hard to diagnose and usually indicate a 
hardware problem.
Please, give us more information about your laptop. You can use "inxi" 
utility for that:

    # inxi -a -v8 -z -za

You can send the output from "inxi" to paste service[1] and provide us 
with just a link to it in next mail.


Is laptop's battery(-ies) in good shape, i.e. it can sustain laptop at 
least for a few minutes with PSU disconnected?
If battery is dead, disconnect the battery and try to reproduce the 
issue with only PSU connected.



It it not particular warm before and journalctl / dmsg logs are looking
unsuspicious around these sudden shutdowns for me.

Any pointer what else could be investigated (and I do realize that these are
very vague symptoms) would be appreciated!

A good place to start is to check journald logs for previous boot:
    # journalctl --boot -1

"-1" is an index, so "0" is current boot, "-2" is two boots prior, etc.


[1] https://paste.debian.net/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: apcupsd sc420i line voltage

[Alexander V. Makartsev]

On 07.12.2022 19:47, Vukovics Mihály wrote:

Hello Community,

I am strugling with APC SmartUPS SC420I and apcupsd.
The UPS is connected via usb-to serial adapter and 940-0024E cable.

The UPS is recognized as "Back-UPS Pro 280" instead of SC420I and the 
line voltage is displayed as 120V instead of the real 230V.


UPSTYPE apcsmart
DEVICE /dev/ttyUSB0

Has any of  you experienced like this and know how to fix at least the 
line voltage value?


Maybe it would be better to try "nut-server" which could have a better 
apcsmart driver implementation?
I have APC SmartUPS SC1500I with similar setup, USB-to-COM adapter and 
self-made APC UPS Smart cable and it is working fine with nut-server, 
both monitoring and control.


Another thing to check could be an usb-to-serial adapter. I've seen some 
cheap low-quality adapters which were created with counterfeit ICs.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: Questions about installing Debian on a laptop

[Alexander V. Makartsev]

On 04.12.2022 18:52, Gabor Urban wrote:

Hi,

I am planning to install Debian on a laptop the first time.That will 
not be my first installing but I never used notebooks for that. I have 
found a lot of useful information but I would like to have some 
guidance at the start.


What are the most important issues selecting a laptop I should be 
mindful about?


Probably the most problematic point of a Linux laptop is a wireless 
connectivity, both WiFi and Bluetooth.
It's always better to check beforehand if wireless adapter that comes 
preinstalled is supported natively by Linux kernel [1], or would work 
with external drivers (kernel modules) supplied by device manufacturer.
Another thing to keep in mind, is to get a laptop with a MUX switch (AKA 
Advanced Optimus) to be able to select between CPU internal and discrete 
VGAs.
While classic Optimus technology could work out of the box, it is 
usually a headache to make it work, because hardware implementation is 
different for every laptop manufacturer.
So it is better to just select discrete VGA in BIOS as primary, and MUX 
switch allows you to do that.


If you already have the laptop handy you can test how it would work with 
Debian Live [2] before removing preinstalled OS.
It's better to choose from images with non-free firmware included to 
save yourself some time from extra troubleshooting while trying make 
things work.



[1] https://wireless.wiki.kernel.org/en/users/drivers
[2] https://www.debian.org/CD/live/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: firmware-atheros - slow internet

[Alexander V. Makartsev]

On 09.11.2022 05:37, Amn wrote:

I have found impossible to use Bluetooth with my Debian 11 on my :
System Manufacturer    Acer
System Model    Aspire E5-771G
System Type    x64-based PC
System SKU    Aspire E5-771G_0880_1_09
Processor    Intel(R) Core(TM) i5-4210U CPU @ 1.70GHz, 2401 Mhz, 2 
Core(s), 4 Logical Processor(s)

BIOS Version/Date    Insyde Corp. V1.09, 2014-07-17

Though it is an old laptop, I was confident that Debian 11 would be 
able to bring her to its past glory, and it did! with the exception of 
the the Bluetooth capability.


I 'apt install firmware-atheros', but the result were most 
unappreciated since the package, though it did make the Bluetooth 
devices work, it created a problem with the WiFi; the internet was 
super, mega slow.


I still hope to make the Bluetooth work and thus I am posting this 
email in the hope that someone here also had the same experience with 
an Acer Aspire and was able to solve it.


Please note, I am by no means an expert on Linux, I have always use 
M$, but due to my disability I am now unable to purchase a new laptop 
or the new  WinOS. So, go easy on me folks for I am a new comer to the 
*nix world.


Thanks in advance.


We need more information about your system first.
Install necessary utilities first, if they are not already installed:
    $ sudo apt install pciutils usbutils inxi

Send the output from commands below in your next email:
$ lspci -nn -k
$ lsusb --tree -v
$ inxi -E -za
$ journalctl -b --no-pager | grep -iE "wireless|bluetooth|firmware"

You can use pastebin-like service[1] to send long console outputs.
What devices are you trying to connect via bluetooth?



[1] https://paste.debian.net/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: loss of mbmon function

[Alexander V. Makartsev]

On 02.11.2022 05:55, gene heskett wrote:

On 11/1/22 16:52, Alexander V. Makartsev wrote:
First step is to find model and make of IC that provides sensor 
functions.

Does this command gives any clues?
$ sudo sensors-detect


answered yes to all the default NO questions and still got only:
# Chip drivers
coretemp
nct6775

Which are now loaded, but still no voltages or fans are reported.


These two has to be sufficient. Just make sure these modules are loaded:
$ lsmod | grep -iE 'coretemp|nct6775'

Now "sensors" utility should be able to output values from every 
available sources of super I/O IC.

    $ sensors


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: loss of mbmon function

[Alexander V. Makartsev]

On 01.11.2022 18:20, gene heskett wrote:

Greetings all;

I am now suffering from a hang on reboot. And in looking for info, I 
find that gkrellm can only see temps. I don't push this
so they stay in the 29 to 30C range. gkrellm is, and has been part of 
my housekeeping for 20 years.


But mbmon was not installed, but it and all its suggested dependency's 
are now, and two reboots, which took about 20
minutes just to get to the bios screen while dancing a jig on the del 
key. During that time I can hear a
very faint clicking sound from time to time. zero activity on any 
drive controller led, there are two controllers,
one of course on the mobo, and one that interfaces a 4 drive raid10 
for the /home.


Mobo is: Asus PRIME Z370-A II, BIOS 0801 04/24/2019

mbmon claims to run by itself but needs root, and when ran with sudo, 
reports

gene@coyote:~$ sudo mbmon
[sudo] password for gene:
No Hardware Monitor found!!
InitMBInfo: Success

What do you suggest I install so this Asus mobo  can be monitored.

Thanks all. Take care and stay well.

Cheers, Gene Heskett.

First step is to find model and make of IC that provides sensor functions.
Does this command gives any clues?
$ sudo sensors-detect



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: Failure in fuse3 hook prevents initramfs update

[Alexander V. Makartsev]

On 29.10.2022 21:32, Joe Pfeiffer wrote:

Further follow-up:  the problem appears to be that something else has
already put mount.fuse3 in the initramfs.  Replacing the failing line
with

copy_exec /sbin/mount.fuse3 /sbin || true

allows me to create the initramfs, and the system boots, but I doubt
it's an optimal solution.

I don't have the same issues with initramfs, but maybe this will help to 
find what script also puts mount.fuse3 into initramfs:

$ grep -iE 'fuse|fuse3' /usr/share/initramfs-tools/hooks/*

This is the output on my system:
/usr/share/initramfs-tools/hooks/fuse:copy_exec /sbin/mount.fuse3 /sbin
/usr/share/initramfs-tools/hooks/fuse:manual_add_modules fuse
/usr/share/initramfs-tools/hooks/ntfs_3g:PREREQ="fuse"

Also, this will list all fuse-related installed packages:
$ dpkg -l | grep -iE 'fuse|fuse3'

This is the output on my system:
ii  exfat-fuse 1.3.0-2    amd64    read 
and write exFAT driver for FUSE
ii  fuse3 3.10.3-2   amd64    Filesystem 
in Userspace (3.x version)
ii  fuseiso 20070708-3.2+b1    amd64    FUSE 
module to mount ISO filesystem images
ii  libfuse2:amd64 2.9.9-5    amd64    
Filesystem in Userspace (library)
ii  libfuse3-3:amd64 3.10.3-2   amd64    
Filesystem in Userspace (library) (3.x version)
ii  libntfs-3g883 1:2017.3.23AR.3-4+deb11u2  amd64    
read/write NTFS driver for FUSE (runtime library)
ii  lxcfs 4.0.7-1    amd64    FUSE based 
filesystem for LXC
ii  ntfs-3g 1:2017.3.23AR.3-4+deb11u2  amd64    
read/write NTFS driver for FUSE


If you've got another script in hooks directory that also tries to copy 
mount.fuse3, you can find a package it comes from:

E.g. $ apt-file find /usr/share/initramfs-tools/hooks/fuse

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: Whom to send bugreport?

[Alexander V. Makartsev]

On 20.10.2022 12:15, Hans wrote:

Dear ladies and gentlemen,

guess there is the following problem:

There is a third-party module for a kernel, which has to be compiled for the
kernel.  This build is working fine!

After upgrading to a newer version of the kernel and of course newer kernel-
headers, the build is failing now with a crash. As the third-party module was
not changed, the failure depends on some changes in the kernel-header files.

Whom the bugreport hast to be sent? The kernel-header maintainers? The third-
party?

IMO third-party should maintain their code and fix the issue.
If third-party would determine that this issue is not their fault, then 
they should report a regression bug upstream, explaining the problem in 
technical detail and/or offer patches.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: Cheap NAS

[Alexander V. Makartsev]

On 14.10.2022 00:23, pa...@quillandmouse.com wrote:

Folks:

This has likely already been covered, but I don't know a way to search
the archives for it.

I'm interested in a desktop NAS with maybe 4 bays, which works with
Linux (and free software), isn't hugely expensive, and is *not* a PC
(you could do a NAS with a PC).

Any suggestions would be helpful.

Paul

If you are looking for a "ready-to-go" NAS appliance, then I recommend 
to look for Synology products.
Their DS value and DS plus series are based on relatively good hardware 
and have stable software/firmware with too many features and options.
2-4 bay variants with 1Gb Ethernet ports are very inexpensive and IMO 
perfect for SOHO.


That is if you are not on an adventure to build your own NAS, of course. :)
And if you are, then prices, for anything not ATX, with small 
form-factors and custom cases, would be too high.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: SSD Optimization - Crucial CT1000MX500SSD1

[Alexander V. Makartsev]

On 29.09.2022 18:54, Marcelo Laia wrote:

Hi,

Recently, I bought a SSD SATA Crucial CT1000MX500SSD1.

Nowadays, one week a go, debian testing system got crashed because 
partition got read only.


It is expected for partitions to fallback into read-only mode if there 
are errors reported from the driver.


I suspect this is a reason from some upgrade and I suspect that an SSD 
optimization can solve this problem.


SSD firmware is up to date. This version is M3CR043.

Based on the info you've sent, there are a few options to try:
1. Replace SATA cable with a known working one. I suggest this because 
there are a few errors were registered in SMART Attribute 199.
2. Test the drive within the most basic conditions, like one GPT 
partition and Ext4 filesystem without "discard" option. Leave extra 
layers like LVM, LUKS, TRIM feature, etc, aside for now.
3. There is a possibility a BIOS\Firmware update for your motherboard 
could improve compatibility with SSD's internal controller.


There are no "SSD optimization" exist that would be required for a SSD 
to function. They are "it just works" kind of devices.
Overall SMART looks clean to me, like the one from a brand new SSD would 
look like.
There are multiple similar reports on the Internet, so I'd suspect it is 
more of a hardware\firmware compatibility issue than a faulty drive.
E.g. there is a high chance this SSD would work just fine inside another 
PC (with different motherboard, ICH, BIOS, etc.)


Can you show us more info about your PC/Laptop? A report from "inxi" 
would be great.
You can get all possible info and filter out private data, with these 
parameters:

    $ sudo inxi -a -v 8 -za



Here is some informations:

:~$ sudo smartctl -a /dev/sda

https://pastebin.com/Jyrhn1A2

:~$ sudo journalctl --since "2022-09-25 00:00:00" | grep sda

https://pastebin.com/QtCqpJPm

:~$ sudo journalctl --since "2022-09-25 00:00:00" | grep ata1

https://pastebin.com/x5QdaYQU

:~$ sudo journalctl --since "2022-09-25 00:00:00" | grep error

https://pastebin.com/1hEPm7YX

Have you some clue and/or advise here?

Thank you so much!




--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Currently on x11vnc, looking for reliable VNC solution?

[Alexander V. Makartsev]

On 07.09.2022 13:41, piorunz wrote:

On 07/09/2022 05:58, notoneofmyseeds wrote:

On 07.09.22 06:19, Alexander V. Makartsev wrote:




I've switched to NoMachine [1] a long time ago.
It has all features I need, which are multi-platform and cross-OS
support, public key authentication, reliable file transfer between
hosts, and completely free no strings attached license for personal 
use.


and the there's anydesk, with conditions just as nomachine.

anydesk.com




[1] https://www.nomachine.com/


Thanks for your replies guys. These solutions are overkill to my needs,
I just need reliable LAN access from one machine to another,

NoMachine does exactly that.


as for WAN
access I already have ssh tunnel which tunnels all traffic I want if
need be. So, I don't think I need external, commercial, not open source
solution for my simple remote access. 
NoMachine is not external service like anydesk or teamviewer. Doesn't 
nag you about license or anything.
It is a completely self-hosted solution and doesn't require Internet 
access to work. You setup your own SSH keys and\or passwords, ports, 
settings, etc.
For me at home NoMachine provides a physical-like-access experience to 
all my Linux and Windows hosts and IMO was a major step forward from 
slow, ancient and insecure VNC.



I'd rather fix VNC server I have
right now, or switch to different VNC server. Anyone has experience with
VNC, or similar LAN protocols, which work? Thanks in advance.

I was in search for a VNC replacement once too. Good luck.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Currently on x11vnc, looking for reliable VNC solution?

[Alexander V. Makartsev]

On 07.09.2022 01:49, piorunz wrote:

Hi all,
...

Any suggestions welcome!


I've switched to NoMachine [1] a long time ago.
It has all features I need, which are multi-platform and cross-OS 
support, public key authentication, reliable file transfer between 
hosts, and completely free no strings attached license for personal use.



[1] https://www.nomachine.com/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: BlueTooth - Acer - Qualcomm Atheros

[Alexander V. Makartsev]

On 26.08.2022 21:24, Amn wrote:


Hey there!!

As the subject suggests, this combo [BlueTooth - Acer - Qualcomm 
Atheros] is my problem.
But before going any further I'd like to leave clear that I am not, by 
any means, a Linux Power User [LPU], I just installed Debian 11 in my 
laptop and are now discovering the pros and cons of this OS, that's it!


After installing Debian 11 from an USB, I found out that the 
blue-tooth wasn't working, so searching the net I found this web pages 
that attempt to fix the issue, among them are



There is pretty good ArchWiki page about Bluetooth setup. [1]
Basically, first you need to install bluetooth stack provided by BlueZ 
and install firmware for your hardware, before you will proceed to the 
other things, like changing PulseAudio config.

To install bluetooth stack:
$ sudo apt install bluetooth bluez bluez-tools bluez-firmware blueman
And to install firmware (from non-free section [2] ):
  $ sudo apt install firmware-atheros

To check if there is a Bluetooth adapter:
    $ hciconfig -a

If there is an adapter, you can use GUI application "Bluetooth Manager" 
for an additional setup and BT device pairing.



You can also install additional packages that will help to gather 
information about your hardware:

  $ sudo apt install lshw pciutils usbutils inxi

They will provide "lshw", "lspci", "lsusb", "inxi" console utilities.


[1] https://wiki.archlinux.org/title/bluetooth
[2] https://wiki.debian.org/SourcesList

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: closing Bullseye bugs pointing to a fix in Unstable?

[Alexander V. Makartsev]

On 23.07.2022 12:39, Thomas Schmitt wrote:

Hi,


Surely "Closes:" is very convenient, but wouldn't you agree that this
puts the users of Stable at a disadvantage?

Well, "stable" means old software with old bugs. Those who want the new
bugs, which are introduced by fixing the old ones, have to run something
else.
Then why "nvidia-driver" in Stable was switched from previous 
"460.91.03-1" version to "470.129.06-6~deb11u1"?

    $ rmadison nvidia-driver
    nvidia-driver | 340.106-1    | 
oldoldoldstable/non-free    | amd64, armhf, i386
    nvidia-driver | 390.138-1    | 
oldoldstable/non-free   | amd64, armhf, i386
    nvidia-driver | 418.152.00-1~bpo9+1  | 
stretch-backports/non-free  | amd64, armhf, i386
    nvidia-driver | 418.211.00-1 | 
oldstable/non-free  | amd64, armhf, i386
    nvidia-driver | 470.103.01-1~bpo11+1 | 
bullseye-backports/non-free | amd64, arm64
    nvidia-driver | 470.129.06-6~deb11u1~bpo10+1 | 
buster-backports/non-free   | amd64, arm64
    nvidia-driver | 470.129.06-6~deb11u1 | 
stable/non-free | amd64, arm64
    nvidia-driver | 470.129.06-6 | 
testing/non-free    | amd64, arm64
    nvidia-driver | 470.129.06-6 | 
unstable/non-free   | amd64, arm64
    nvidia-driver | 510.73.08-3  | 
experimental/non-free   | amd64, arm64


This change made my system unstable, causing it to freeze sporadically 
with "Xid" error in syslog.
    kernel: NVRM: GPU at PCI::01:00: 
GPU-7565947f-d476-7159-d106-172c793521e6

    kernel: NVRM: Xid (PCI::01:00): 8, pid=2039, Channel 0010

And now I can't perform an easy enough downgrade to last working version.
Freezes are only happening with 470.xxx versions of the "nvidia-driver", 
every other versions up to 460.91.03 worked flawlessly.


    $ lspci
    ...
    01:00.0 VGA compatible controller: NVIDIA Corporation GP106 
[GeForce GTX 1060 6GB] (rev a1) (prog-if 00 [VGA controller])

    Subsystem: NVIDIA Corporation GP106 [GeForce GTX 1060 6GB]
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- 
ParErr- Stepping- SERR- FastB2B- DisINTx+
    Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast 
>TAbort- SERR- 
    Latency: 0
    Interrupt: pin A routed to IRQ 135
    IOMMU group: 1
    Region 0: Memory at ee00 (32-bit, non-prefetchable) 
[size=16M]

    Region 1: Memory at d000 (64-bit, prefetchable) [size=256M]
    Region 3: Memory at e000 (64-bit, prefetchable) [size=32M]
    Region 5: I/O ports at e000 [size=128]
    Expansion ROM at 000c [virtual] [disabled] [size=128K]
    Capabilities: 
    Kernel driver in use: nvidia
    Kernel modules: nvidia

    $ uname -a
    Linux fortune 5.10.0-16-amd64 #1 SMP Debian 5.10.127-1 (2022-06-30) 
x86_64 GNU/Linux


Should I file a new bug?
If so, what is the best way to do it, if the last freeze happened 4 days 
ago, according by timestamps in syslog, and now I plan to downgrade the 
driver to be able to use full capabilities of my PC?



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: MAC address of bridge interfaces

[Alexander V. Makartsev]

On 29.06.2022 07:27, Steve Keller wrote:

I upgraded a Debian machine from stretch to bullseye and see
a change of the IP address of a ethernet bridge interface.

The bridge has a physical LAN interface as one fixed bridge port
and additional ports for kvm virtual machines I may start.

Before the upgrade the bridge interface got its MAC address from
the physical LAN interface.  With Bullseye this is no longer the
case.  Instead, the MAC address seems to be generated "somehow".
At least, this new address is fixed and doesn't change on each
reboot.

But how is that MAC address generated?  Is it stored somewhere?
Can I set it to an address of my own preference?
And why was this changed, why don't we still use the address of
the physical port connected to it?

Steve

I didn't experienced any problems with libvirt networks after host OS 
dist-upgrades.
On my KVM host, physical ethernet interface and virtual network bridges 
(route mode) have different MAC addresses.
Libvirt\KVM network configurations are explained in the libvirt 
documentation¹.
I don't remember if I ever had to assign any MAC addresses for virtual 
bridges or ports manually.
I've used "Virtual Machine Manager" to create virtual networks and used 
auto-generated MAC addresses, as suggested in the docs.
At some point in time I had to make use of fixed DHCP IP addresses for 
my VMs and also made hook scripts to dynamically add\remove firewall 
rules for VM guests.
All virtual network configurations are defined in xml format and "virsh" 
utility is used to view\edit them.

    $ sudo virsh net-list
    $ sudo virsh net-dumpxml --network 
    $ sudo virsh net-port-list --network 
    $ sudo virsh net-port-dumpxml --port  --network network>



¹ : https://libvirt.org/formatnetwork.html#addressing

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: SSD Optimization and tweaks - Looking for tips/recomendations

[Alexander V. Makartsev]

On 28.06.2022 22:25, Marcelo Laia wrote:

Hi,

I bought a SSD solid disk and will perform a fresh install on it. Debian 
testing. I've never used such a disc.

I bought a Crucial CT1000MX500SSD1 (1TB 3D NAND Crucial SATA MX500 Internal SSD 
(with 9.5mm adapter) — 6.35cm (2.5in) and 7mm).

I read the recommendations on the https://wiki.debian.org/SSDOptimization page.

However, I still have some doubts:

1. Use ext4 or LVM partitioning?
One is a filesystem and the other is a partition technology, a 
completely different thing.
My recommendation is to use whatever you are comfortable and experienced 
with.
Unless of course you feel adventurous and want to learn new things and 
use extra features. More about that below.



2. I read in the Warnming section that some discs contain bugs, including Crucial. But I don't know 
if I need to use or not use "discard" on this disk (CT1000MX500SSD1). If I need to 
proceed with use "discard", would you please have any tips on how to do it? I didn't 
understand how to do this.
Most of those warnings are outdated and nowadays all major SSD 
manufacturers have developed stable firmware for their products.
Personally, I don't use "discard" mount option, instead I manually run 
"fstrim" utility once a few months.
But, my SSD serves me as a system disk and that is not a write-intensive 
workload.



3. Should I reserve a swap partition or not? I always had one on hdd disks. I 
was in doubt, too.
Yes, you should. The difference in performance will be noticeable, 
whenever your system would resort to swap usage.



4. Any other recommendations to improve the performance and lifespan of this 
disk?
A performance improvement is a complex task and in some cases could be 
impossible due to hardware limits of the host or the SSD device itself.
On modern system with SATA3 interface, there is not much you can do to 
increase a performance of SSD, other than optimizing your workloads and 
maybe choosing among different flash-optimized filesystems.
I'd recommend to check out a somewhat recent Phoronix article¹ about 
filesystem benchmarks.
And then again it depends on your workload if these optimizations even 
worth doing and the time spent learning a new techs.


So the bottom line is, if your workload is low, then you should stop 
worrying and simply enjoy your new SSD.
It will last for many years and probably will be moved to a newly build 
PC to serve as an additional storage device there.



Thank you!



¹ : 
https://www.phoronix.com/scan.php?page=article=linux-58-filesystems=1


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: disk mount problem

[Alexander V. Makartsev]

On 09.06.2022 23:52, ghe2001 wrote:

Supermicro desktop, Debian bullseye

There are failures when I try to mount a disk.  Fstab:
...
I've never seen the first one fail, I've never tried the second, the 
third and fifth work as expected, and the forth fails.  (On boot, the 
/blackHole line is commented out.)

Any thoughts?  Have I missed something?

You have a typo in options. Should be:

UUID=21dcbfda-3884-404f-855f-693d1efa2f06    /blackHole    ext4 
defaults    0    0



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: debian app on UserLAnd: upgrading to bullseye

[Alexander V. Makartsev]

On 09.06.2022 19:41, agyaana...@yahoo.com wrote:

i have installed debian (buster) on userland app (UserLAnd is an open-source 
app which allows you to run several Linux distributions like Ubuntu,
Debian, and Kali.)

i am keen to upgrade to bullseye.
my action:
* sudo apt-get update && apt-get full-upgrade
* sudo nano /etc/apt/sources.list: replace buster with bullseye
* sudo apt-get update && apt-get upgrade && apt-get dist-upgrade

am i missing on something. i have limited data plan. hence, requesting opinion 
or suggestions here.
In addition to suggestion from Andrew you should skim through Bullseye's 
release notes¹.



[1] https://www.debian.org/releases/stable/amd64/release-notes/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Debian 10 --> 11 on Dell R740: network interfaces renamed

[Alexander V. Makartsev]

On 09.06.2022 00:04, Harald Dunkel wrote:

Hi folks,

after the upgrade to Debian 11 some network interfaces in my
Dell R740 got renamed. Before:

# lshw -class network -short
H/W path  Device  Class  Description

/0/2/0    eno1    network    Ethernet Controller 10G 
X550T
/0/2/0.1  eno2    network    Ethernet Controller 10G 
X550T
/0/3/0    eno3    network    I350 Gigabit Network 
Connection
/0/3/0.1  eno4    network    I350 Gigabit Network 
Connection
/0/103/0  ens3f0  network    I350 Gigabit Network 
Connection
/0/103/0.1    ens3f1  network    I350 Gigabit Network 
Connection



After:

# lshw -class network -short
H/W path  Device Class  Description
===
/0/2/0    eno1   network    Ethernet Controller 10G X550T
/0/2/0.1  eno2   network    Ethernet Controller 10G X550T
/0/3/0    eno3   network    I350 Gigabit Network 
Connection
/0/3/0.1  eno4   network    I350 Gigabit Network 
Connection
/0/103/0  enp94s0f0  network    I350 Gigabit Network 
Connection
/0/103/0.1    enp94s0f1  network    I350 Gigabit Network 
Connection



How comes? AFAIR these predictable interface names had been introduced
to get *stable* names, if the hardware is changed. Since it is more
likely to get a kernel upgrade than new network hardware I wonder of the
predictable names could be made even more predictable?

For me the best solution is to manually create .link¹ files for every 
interface, to name them to something nice, like "ether0" or "wan0".


[1] https://www.freedesktop.org/software/systemd/man/systemd.link.html

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: firefox misbehaviour

[Alexander V. Makartsev]

On 21.05.2022 17:33, Hans wrote:

Hmm, this combination is new to me. As I am using plasma5, there is
, which changes the mousepointer to a skull.

Clicking then with the skull on some window, it will kill this window/
application. However, this does not work.
  
Pressing  should switch to console 1, (F2 to 2, F3 to 3 and so

on) but this does not work, too.

As I said, completely frozen! Oh, and it is also not possible, to login from
another computer (i.e. with ssh), as due to the freeze authentication and
other things do not work then.
Last week, when my system froze, I was able to use SysRq¹ key 
combinations to sync disks, remount and reboot.
Might be a good idea to check if it is enabled² (it should be enabled by 
default AFAICR), and keep instructions handy, if your system will freeze 
once again.



¹ https://en.wikipedia.org/wiki/Magic_SysRq_key
² 
https://www.debian.org/doc/manuals/securing-debian-manual/restrict-sysrq.en.html


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: firefox misbehaviour

[Alexander V. Makartsev]

On 21.05.2022 10:57, Hans wrote:

Am Samstag, 21. Mai 2022, 05:09:38 CEST schrieb Russell L. Harris:
Hi Russell,

I do not have those issues you describe, but on my system firefox sometimes
completely freezes the system when calling some special websites.

Then nothing can be done, only the mouse is moving but no mouse clicking, nor
keyboard input is possible then.

The only way is the hard way: long press power button, so it switches off.

I admit, there are some plugins installed, like noscript and ublock origin,
which also might interfere.

However, for testing purposes on those sites I temporarly deactivate the
plugins.

And of course, it can not be excluded, that these sites also interfere with
the graphics driver, which is Nvidia.
I have had similar issues with system locking up on rare occasion, 
definitely noticeable because this system was rock solid stable for years.
For me it began to happen when I've updated "nvidia-driver" to version 
"470.103" from "proposed-updates".
There is also XID #8 error¹ in syslog, at the time when a freeze occurs, 
suggesting it is a Nvidia driver problem.
To verify my suspicion, I've downgraded "nvidia-driver" to version 
"460.91" from "stable" and freezes are gone.
I didn't file a bug report about this, in fact I don't know if I should 
and where to report it, but my system is running for a week now without 
any problems.



¹ https://docs.nvidia.com/deploy/xid-errors/index.html

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Crucial SSDs and Debian Bullseye

[Alexander V. Makartsev]

On 27.04.2022 20:37, Tom Browder wrote:

...
If either of those fail to see it, I’m afraid I toasted it. I don’t 
think that will qualify for a return.

...


What makes you think that way?
It is next to impossible to actually mishandle and "toast" a device 
simply by unpacking it and connecting to a SATA port. (Even when PC is 
powered on¹).
SSDs are not fragile, they are electrically compatible with SATA 
standard, so both data and power port parts won't do any damage to a 
device, no matter how old or new the hardware is.
If you won't manage to get it working with another PC, then you simply 
got a faulty product and have a warranty to get it replaced.
I've seen a fair share of faulty brand new devices in my life, memory 
sticks, motherboards, HDDs.
This is rare nowadays and I don't handle large volumes of PC parts 
anymore, but it still happen here and there.



¹ It is still a risk and is not recommended at all. It is always safer 
to do any work with powered off computer and disconnected battery if it 
is a laptop.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Crucial SSDs and Debian Bullseye

[Alexander V. Makartsev]

On 27.04.2022 17:17, Tom Browder wrote:
On Wed, Apr 27, 2022 at 07:08 Alexander V. Makartsev 
 wrote:

...

The laptop is a Toshiba C655D-S5136 Satellite.
The SSD is a Crucial MX500 1000GB 2.5 inch SSD.

Thanks, Alexander.

-Tom

Crucial MX500 is based on SM2258 controller IC from Silicon Motion.
In general, I usually select SSDs based on controller ICs from Phison. 
They are in my experience have better backward-compatibility with older 
hardware.
The make\brand of the SSD itself doesn't really matter, if you choose 
appropriately the rest of SSD parts (i.e. NAND type, bits per cell, etc)


There is an one workaround you should try:
Inside BIOS of the laptop, check if there is an option to switch between 
protocols in HDD or Chipset sections. They could be called "Legacy", 
"Native" and "AHCI".

Make sure to set it to "AHCI", which should be default for any SSDs.
Some SSDs will still work in "Legacy" mode, but it could cause a 
compatibility issues.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄

Re: Crucial SSDs and Debian Bullseye

[Alexander V. Makartsev]

On 27.04.2022 16:06, Tom Browder wrote:
I am trying to replace the original hard drive on an old Toshiba 
laptop with a 1 TB SSD from Crucial. (I had recently successfully done 
that in an old Dell Latitude and had no problems.)


I first did a clean install of Debian 11 on the old drive to ensure 
the laptop works okay. Then I installed the new SSD and it can't find 
the drive. From what I can find at Crucial, I need to install their 
Storage Executive program on a Windows host, look up the SSD to a 
USB/SATA connector on that host, and configure or install the firmware 
onto the SSD.


I've never heard anything like that and I've worked with many 
consumer-grade SSDs.
Usually all SSDs "just work". They may come pre-partitioned and 
pre-formatted, but this could be reconfigured with any standard utility 
programs.
The only thing I can think of, is that it could require usage of some 
vendor-specific proprietary software to setup hardware encryption and/or

to update currently flashed firmware to newer versions.

I think, you've encountered a hardware compatibility issue between an 
old ICH controller on the host and a SATA controller on the SSD drive.
Some Samsung SSDs with their custom drive controller ICs were affected 
by this issue, refusing to work with some older chipsets and at lower 
SATA speeds.
I suspect it could be done on purpose, because vendor doesn't want to 
look bad when people benchmark their new drive and post sub-par results 
online.


Can you provide exact model\make of your host and the SSD drive?

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Copy/paste between host and KVM/windows 10 guest??

[Alexander V. Makartsev]

On 08.04.2022 14:23, didier gaumet wrote:

Le vendredi 08 avril 2022 à 13:48 +0500, Alexander V. Makartsev a
écrit :

I've had "virtio-win-guest-tools" installed inside Win10 guest

does it appear in the list of the Windows installed applications list
(Windows parameters menu)?
I just verified: that's the only thing I installed in the Win10 guest
related to guest drivers/utilities. (Bullseye here)

Yes it does.

After some investigations, I've managed to get copy\paste action working.
So, to answer Dennis' question, this is a list of things necessary for 
it to work:

1. "virtio-win-guest-tools" has to be installed inside Windows guest.
2. VM guest configuration has to have "Channel spice" device (type 
"spicevmc" with default settings)
3. Required services called "QEMU Guest Agent" and "Spice Agent" has to 
be in running state inside Windows guest.


Adding user to groups "libvirt" and\or "libvirt-qemu" is not necessary 
to get copy\paste functions to work.
And it is probably more secure not to add user to these groups, 
especially if it is your PC and you are the only user.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Copy/paste between host and KVM/windows 10 guest??

[Alexander V. Makartsev]

On 08.04.2022 13:19, didier gaumet wrote:

Hello,

RHEL/Fedora provide guest (virtio) drivers and agents. I have installed
the whole bunch (ISO) into a Win10 KVM guest and the copy/paste is
working properly but I think what is required here is only the Spîce
guest agent.
https://github.com/virtio-win/virtio-win-pkg-scripts/blob/master/README.md
(just run the .exe installer (at the root of the iso) in the guest)
I've had "virtio-win-guest-tools" installed inside Win10 guest from the 
very beginning, but copy\paste actions from host to guest and vise versa 
still don't work.
There must be something else to setup, perhaps clipboard becomes 
unavailable due to user switching?

    Ex. "sudo virsh list --all" works, but "virsh list --all" doesn't.
And when I run "Virtual Machine Manager" it asks me for password.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Copy/paste between host and KVM/windows 10 guest??

[Alexander V. Makartsev]

On 08.04.2022 04:18, Dennis Wicks wrote:
It doesn't work for me and the suggested solutions I have found are 
referring to a linux guest.


TIA for pointers, suggestions, solutions!
Denniis

I've never bothered to make copy\paste actions work natively (if that is 
even possible) and always used RDP to connect to Windows guest instead 
of SPICE.
By using RDP connection you solve many quality of life issues, like 
copy\paste, file sharing, smoother GUI performance, etc.

I use "Remmina" as a front end, to manage remote desktop connections.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Install BIOVIA_2021.DS2021client

[Alexander V. Makartsev]

On 28.03.2022 23:33, Stephen P. Molnar wrote:

Has anyone managed the installation of the BIOVIA_2021.DS2021client?

Thanks in advance
-- Stephen P. Molnar, Ph.D. www.molecular-modeling.net 614.312.7528 
(c) Skype: smolnar1
Out of pure curiosity, I've managed to run this proprietary application¹ 
on Debian 11, but to make it happen I had to manually unpack and 
"install" it, and also implement a few hacks.
Installer scripts don't work and were poorly tested, also this software 
depends on outdated "libpng15.so.15" dynamic library which is not in 
Debian repos anymore.
So it is possible to run it, but there is no guarantee how this ugly 
mess of a software will work after manual "installation" and I suggest 
you to ask for updated version from the developers.
It makes sense, at least because they ask money for a license, even 
though AFAICS it could be used as "free" viewer/editor.
There are also community forums² exist for this application, so you 
probably better to ask for assistance there.



¹ https://imgur.com/a/zBoDwRO
² https://www.3ds.com/products-services/biovia/communities/

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Thunderbird security

[Alexander V. Makartsev]

On 26.03.2022 13:50, André Rodier wrote:

Hi all,

I would like to collect, from this thread, your experience and opinion 
about Mozilla Thunderbird, in term of security.


I am registered on The Debian security list, and I see a lot of CVE 
coming, some of them with a high score, mentioning execution of 
arbitrary code or information disclosure.


Most of them seems pretty severe to me, and I am now running 
Thunderbird in firejail. However, I wonder if such vulnerability would 
allow a remote attacker to send an email, and get, for instance, the 
credentials stored in Thunderbird, with or without master password.


This seem habitual to me, compared to other mail clients in Debian, 
like evolution / claws, etc...


In term of security, Which email clients, or which practices, you 
would recommend to me ?


Thanks for your understanding and advice, but please, I don't want to 
start a troll.


I've used Thunderbird for many years on different platforms. It is my 
favorite mail client and I've never had any major or security problems 
with it.
When it comes to security, it is a good thing to have a healthy dose of 
paranoia and monitor most recent known threats and vulnerabilities,
however the actual exploitation of them is usually quite difficult if 
not impossible, especially if you keep your software up-to-date.


When I search for CVEs for a current version of Thunderbird:
    https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=Thunderbird+91
I don't see any results that could affect 91 version. All of them are 
for older ( < 91 ) versions of Thunderbird.


There is always a possibility of some 0-day vulnerability in any 
software, so if you being smart and exercise some precaution procedures 
you still could be fine.

There are many ways, ex.:
You can disable JavaScript in Thunderbird altogether using 
"about:config" page.
Never open any URLs inside Thunderbird and copy-paste and edit them 
instead, because many of them crafted for purpose of tracking.
Don't open any attachments right away, but save them to disk and inspect 
them instead, especially if they come from unknown sources.
Also, any exploit that could be received by mail has to pass through 
many filters and AV scanners before it will be delivered, so it makes 
exploitation of known vulnerabilities even more difficult for the badguys.
Protecting you credentials with Master Password is a good way to protect 
your data if credential db files were somehow stolen by data-miner class 
malware, completely unrelated to Thunderbird.


Best antivirus is your head and healthy work habits.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: PCIe 3.0 8 port SATA 6 Gbps HBA for SOHO network

[Alexander V. Makartsev]

On 18.03.2022 08:18, David Christensen wrote:


STFW I am unable to determine if the LSI 9207-8i HBA (SAS 2308 I/O 
controller) is supported by Linux and Debian (?).  Suggestions?

This adapter has VEN_ID 1000 & DEV_ID 0087.¹
Output from "$ sudo modinfo mpt3sas" suggests that devices based on this 
controller should be supported.



¹ https://pci-ids.ucw.cz/read/PC/1000/0087

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: PCIe 3.0 8 port SATA 6 Gbps HBA for SOHO network

[Alexander V. Makartsev]

On 18.03.2022 04:59, David Christensen wrote:

debian-user:

I have a SOHO network with various x86_64 servers/ workstations/ 
desktops (FreeBSD, Debian, Windows) and an assortment of phones/ pads/ 
pods/ televisions/ game consoles, etc..  I would like to add more SATA 
6 Gbps HDD's and/or SSD's to the x86_64 machines and am looking for a 
suitable PCIe 3.0 8 port SATA 6 Gbps host bus adapter (non-RAID).



The LSI® SAS 9207-8i PCI Express® to 6Gb/s Serial Attached SCSI (SAS)
Host Bus Adapter looks appealing:

https://docs.broadcom.com/doc/12353331
In my opinion Broadcom\LSI\Avago is one of the best hardware storage 
solutions you can get.

But, this model is EOL.
I mean if it works it works, but in case of any problems with it in the 
future, there will be no official support, no updates to firmware, no 
updates/fixes for drivers.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: recommend music player?

[Alexander V. Makartsev]

On 16.03.2022 22:54, kaye n wrote:

Hello Friends!

I am currently using Debian GNU/Linux 11 (bullseye)

Can anyone recommend a good music player with an equalizer where I can 
choose Pop, Rock, etc.


Thank you!

That could be "Qmmp". It looks and function like an old-school WinAmp.

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: linux kernel and nvidia - never ending story

[Alexander V. Makartsev]

On 08.03.2022 14:40, Hans wrote:

Am Dienstag, 8. März 2022, 00:07:05 CET schrieb Alexander V. Makartsev:
Yes, I am sure, no other version is working. It is GeForce G210 and GeForce
G86m, both NEED 340xx.
In that case, I've never actually tried to do it, but since Debian 
support for 340xx version is officially EoL, I think it's ok to install 
officially supported driver¹ from Nvidia.


Or, if you feel adventurous, you can build a backported packages of 
nvidia legacy 340xx driver from buster-backports.
I've always used parts of this guide² and had successfully build 
packages from source³ package and DKMS module was also build 
successfully during package installation.

You can get source package using "dget":
    $ dget 
http://deb.debian.org/debian/pool/non-free/n/nvidia-graphics-drivers-legacy-340xx/nvidia-graphics-drivers-legacy-340xx_340.108-10~bpo10+1.dsc


No code patches were made by me, except changing "changelog" file via 
'dch' and bumping up version numbers.


    $ uname -a
Linux hostname 5.10.0-11-amd64 #1 SMP Debian 5.10.92-1 (2022-01-18) 
x86_64 GNU/Linux


=8<=8<=8<=8<=
...
Setting up nvidia-legacy-340xx-kernel-dkms (340.108-10~bpo11+1) ...
Loading new nvidia-legacy-340xx-340.108 DKMS files...
Building for 5.10.0-11-amd64
Building initial module for 5.10.0-11-amd64
Done.

nvidia-legacy-340xx.ko:
Running module version sanity check.
 - Original module
   - No original module exists within this kernel
 - Installation
   - Installing to /lib/modules/5.10.0-11-amd64/updates/dkms/

nvidia-legacy-340xx-uvm.ko:
Running module version sanity check.
 - Original module
   - No original module exists within this kernel
 - Installation
   - Installing to /lib/modules/5.10.0-11-amd64/updates/dkms/

depmod

DKMS: install completed.
...
=8<=8<=8<=8<=

Here is listing of all build packages for amd64 arch:
=8<=8<=8<=8<=
$ ls -1
libegl1-nvidia-legacy-340xx_340.108-10~bpo11+1_amd64.deb
libgl1-nvidia-legacy-340xx-glx_340.108-10~bpo11+1_amd64.deb
libgles1-nvidia-legacy-340xx_340.108-10~bpo11+1_amd64.deb
libgles2-nvidia-legacy-340xx_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-cfg1_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-compiler_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-cuda1_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-eglcore_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-encode1_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-fbc1_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-glcore_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-ifr1_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-ml1_340.108-10~bpo11+1_amd64.deb
libnvidia-legacy-340xx-nvcuvid1_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-alternative_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-driver_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-driver-bin_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-driver-libs_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-kernel-dkms_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-kernel-source_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-kernel-support_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-opencl-icd_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-smi_340.108-10~bpo11+1_amd64.deb
nvidia-legacy-340xx-vdpau-driver_340.108-10~bpo11+1_amd64.deb
xserver-xorg-video-nvidia-legacy-340xx_340.108-10~bpo11+1_amd64.deb
=8<=8<=8<=8<=

Same listing is for i386 arch too.
To build 32-bit packages the process essentially the same, but you 
(AFAIK still) have to do it under 32-bit environment (chroot, lxc/lxd 
container, VM, etc).


I didn't tested this build on actual hardware yet, but it looks 
promising to me.

Good luck.


¹ https://www.nvidia.co.uk/Download/driverResults.aspx/156193/en-uk
² https://wiki.debian.org/SimpleBackportCreation
³ 
https://packages.debian.org/buster-backports/nvidia-legacy-340xx-kernel-dkms


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: linux kernel and nvidia - never ending story

[Alexander V. Makartsev]

On 07.03.2022 23:49, Hans wrote:

Dear list,

how find the correct words, without being upset or stepping on someones feet.
But I believe, debian hates Nvidia, and debian does not want, to use Nvidia.

I am now for a long time using debian and also using nvidia graphic cards for
almost the same long time.

But whenever debian ships a new kernel version, the proprietrary nvidia kernel
modules can not be built. If lucky, there is a patch for it after months.

Yes, modern Nvidia cards are supported, but using an older notebook you can
not change the graphics card.
...
I have several older notebooks, that my customers use. They worked perfectly
with the proprietrary driver from Nvidia. But after update to bullseye, it was
hardly get them running again. And why? They have an old graphics card in
their notebooks, and they use Nvidia cards, specially the legacy 340xx.
Out of curiosity, can you name a model of nvidia GPU(-s) you have 
trouble with?
Are you sure you need legacy 340xx driver specifically? That version is 
for hardware that was released in 2009-2010 and older.

I think there is a chance you can install legacy 390xx version instead.
You can use 'nvidia-detect' program to check your hardware and get a 
driver version recommendation.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: XFCE: ALT-F1 shows the wrong menu on Debian 11

[Alexander V. Makartsev]

On 28.02.2022 02:32, José Luis González wrote:

Hi,

Upon upgrading to Debian 11, the ALT+F1 key, which is assigned as a
shortcut to xfce4-popup-applicationsmenu, according to XFCE's settings,
no longer shows the applications menu and instead the app menu button
on my panel appears pressed without the menu unfolded. A second key
press shows the CTRL-ESC (xfdesktop --menu) menu, not the app menu.

What's going on? I'm suffering from the same in both machines I have
with Debian 11.
It looks like "xfce4-popup-applicationsmenu" is a simple wrapper Bash 
script.
All it does is trying to execute "xfce4-panel" with a long line of 
parameters and if that fails, fallback to execute "xfdesktop --menu" 
instead.
Man page for "xfce4-panel" doesn't explain anything about 
"--plugin-event" parameter, so it looks like it was reworked somehow.
The problem is that "xfce4-panel" never fails to execute with that 
undocumented parameter, so fallback option doesn't work.
You can file a bug report about this, or change a command, assigned to 
'Alt+F1' key combination, to

    "/usr/bin/xfce4-panel --plugin-event=applicationsmenu:popup"

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Which nvidia driver? xset +dpms issues. Is my video card dead/broken?

[Alexander V. Makartsev]

On 27.02.2022 04:57, R. Ramesh wrote:

p  nvidia-driver - NVIDIA metapackage
p  nvidia-driver-bin - NVIDIA driver support binaries
p  nvidia-driver-libs - NVIDIA metapackage (OpenGL/GLX/EGL/GLES 
libraries)
p  nvidia-driver-libs:i386 - NVIDIA metapackage (OpenGL/GLX/EGL/GLES 
libraries)
p  nvidia-vdpau-driver:i386 - Video Decode and Presentation API for 
Unix - NVIDIA driver


Do I have a better driver that I should try? Is my video card gone bad?

Your graphics adapter is modern enough so you should install 
"nvidia-driver" package.
Additional essential packages are: "nvidia-driver-bin", 
"nvidia-driver-libs", "nvidia-kernel-support", "nvidia-kernel-dkms", and 
others.


If all of them already installed, try to re-install "nvidia-kernel-dkms" 
package.

It will rebuild DKMS module using kernel sources to match kernel in use.
It is a good idea to do this, after every update of kernel image, to 
make sure a new kernel and DKMS module will be 100% compatible.



--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Hibernate on a Lenovo IdeaPad Yoga 13: XFCE problem?

[Alexander V. Makartsev]

On 26.02.2022 02:23, Charles Curley wrote:

The appears to not offer hibernate. Neither hibernate nor hybrid sleep
are available in the logout menu. Calling an XFCE command to hibernate
("xfce4-session-logout --hibernate") does nothing. Suspend is available
and works.

On installation, I selected the default disk layout and got a swap space
of a measly 976M. I speculated that this was not enough (the machine has
3.7Gi of physical RAM), so I added an 8 Gi swap file.
Creating a swap file won't do. It has to be a swap partition, big enough 
to store memory footprint of a working OS and everything that is 
currently running on it.
Additionally check "initramfs" config file: 
"/etc/initramfs-tools/conf.d/resume"

It needs to have UUID of a swap partition that will be used for hibernation.
Example config:
    $ cat /etc/initramfs-tools/conf.d/resume
    RESUME=UUID=d0331ef1-bc12-473f-bd4a-b4edf0d4d0a8

    $ sudo blkid | grep swap
    /dev/sdb2: LABEL="swap" UUID="d0331ef1-bc12-473f-bd4a-b4edf0d4d0a8" 
TYPE="swap" PARTUUID="88839920-fc62-4f3a-aa7c-e97cd1ed4d5d"


But your case is different, because your swap partition is encrypted. 
I've found an article which looks detailed enough, so give it a shot:

https://help.ubuntu.com/community/EnableHibernateWithEncryptedSwap

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Why is systemd starting Firefox?

[Alexander V. Makartsev]

On 23.01.2022 23:36, local10 wrote:

Hi,

Based on the log records (see below) it looks like systemd is regularly 
starting Firefox. I don't recall setting up anything like that, so why is 
systemd doing it?

Thanks

# cat /var/log/syslog
...
Jan 23 06:31:37 test systemd[922]: Started Firefox ESR - Web Browser.
...
Jan 23 06:31:53 test systemd[922]: 
app-firefox\x2desr-bcb63a3a10404a7fa8443768412b385a.scope: Succeeded.
Jan 23 06:31:53 test systemd[922]: 
app-firefox\x2desr-bcb63a3a10404a7fa8443768412b385a.scope: Consumed 4.746s CPU 
time.
Jan 23 08:05:28 test systemd[922]: 
app-firefox\x2desr-512a87e34cb047e0ace37d7988d04f1b.scope: Succeeded.
Jan 23 08:05:28 test systemd[922]: 
app-firefox\x2desr-512a87e34cb047e0ace37d7988d04f1b.scope: Consumed 2h 43min 
45.540s CPU time.
Jan 23 08:05:31 test systemd[922]: Started Firefox ESR - Web Browser.
...
Jan 23 09:42:17 test dbus-daemon[430]: [system] Activating via systemd: service 
name='org.freedesktop.hostname1' unit='dbus-org.freedesktop.hostname1.service' requested 
by ':1.829' (uid=1007 pid=53514 comm="/usr/lib/firefox-esr/firefox-esr 
-ProfileManager ")
Jan 23 09:43:38 test dbus-daemon[430]: [system] Activating via systemd: service 
name='org.freedesktop.hostname1' unit='dbus-org.freedesktop.hostname1.service' requested 
by ':1.833' (uid=1007 pid=53514 comm="/usr/lib/firefox-esr/firefox-esr 
-ProfileManager ")
...

# systemctl list-units | grep -i firefox


# cat /etc/debian_version && uname -a
11.2
Linux srv07 5.10.0-11-amd64 #1 SMP Debian 5.10.92-1 (2022-01-18) x86_64 
GNU/Linux

firefox-esr is installed on my system (Debian 11 with Xfce) as a dependency.
    $ grep -i "firefox" /var/log/syslog
Returns nothing on my system, but I don't use KDE.

Does this command show anything useful when 'firefox-esr' is running?
    $ systemctl --user status

Systemd's manual describes scopes¹ as special units, so for now I think 
this is how KDE runs and manages applications.

It sure looks suspicious in the logs.


¹ https://www.freedesktop.org/software/systemd/man/systemd.scope.html

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: changed laptop disk -> PXE-E61 & PXE-M0F errors / BootDevice Not Found

[Alexander V. Makartsev]

On 18.01.2022 03:44, Vincent Lefevre wrote:

I tried to change my laptop disk following the instructions found at
https://askubuntu.com/questions/260777/how-to-migrate-an-encrypted-lvm-install-to-a-new-disk

I'm not sure I did everything right as there are some mistakes and
not everything is clear.

In particular, with cfdisk, I initially chose the gpt label type
because this was the default, so that I thought that this was OK.
But "grub-install /dev/sdb" complained ("this GPT partition label
contains no BIOS Boot Partition; embedding won't be possible.
[...]"). So I changed it to MBR with gdisk as explained on
https://unix.stackexchange.com/questions/61142/remove-gpt-default-back-to-mbr

Note: /dev/sdb corresponds to the new disk connected via a SATA-USB
adapter for the transfer.

Then "cfdisk /dev/sdb" showed:

  Disk: /dev/sdb
Size: 931.51 GiB, 1000204886016 bytes, 1953525168 sectors
Label: dos, identifier: 0x

 Device   Boot   Start  End Sectors   Size  Id Type
 /dev/sdb12048  1050623 1048576   512M  83 Linux
 /dev/sdb2 1050624   1953525134  1952474511   931G  8e Linux LVM

This ended up with the following:

# grub-install /dev/sdb
Installing for i386-pc platform.
Installation finished. No error reported.

# update-grub
Generating grub configuration file ...
Found background image: /usr/share/images/desktop-base/desktop-grub.png
Found linux image: /boot/vmlinuz-5.15.0-2-amd64
Found initrd image: /boot/initrd.img-5.15.0-2-amd64
Found linux image: /boot/vmlinuz-5.15.0-1-amd64
Found initrd image: /boot/initrd.img-5.15.0-1-amd64
Warning: os-prober will not be executed to detect other bootable partitions.
Systems on them will not be added to the GRUB boot configuration.
Check GRUB_DISABLE_OS_PROBER documentation entry.
done

Then I unmounted everything related to /dev/sdb.

But after changing the disk and booted, instead of getting GRUB,
I got PXE-E61 and PXE-M0F errors, and also


BootDevice Not Found

Please install an operating system on your hard disk.

Hard Disk - (3F0)

F2 System Diagnostics

For more information, please visit: www.hp.com\go\techcenter\startup


Note that the disk itself is recognized since the system allowed me
to do a quick check on it (no errors).

So, what could be wrong?

It looks like you've forgot to mark boot partition (/dev/sdb1 ?) as 
'Active'.
It is necessary to do for DOS\MBR-style partition table, so stage-2 
bootloader could be found during boot process.


You've got PXE-* errors because BIOS tried to boot from network using 
PXE\BOOTP and failed.


--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: AMD EPYC throttled to 400 mhz

[Alexander V. Makartsev]

On 17.01.2022 18:40, Simon Kainz wrote:


Am 17.01.22 um 11:36 schrieb Alexander V. Makartsev:

On 17.01.2022 14:41, Simon Kainz wrote:

Hello,

we are experiencing spontaneous CPU speed throttlings.

System is a Lenovo  ThinkSystem SR645 with 2
AMD EPYC 7452 32-Core Processor, running

Linux node3 5.10.0-10-amd64 #1 SMP Debian 5.10.84-1 (2021-12-08) x86_64
GNU/Linux

After some time (hours, day, weeks even) the system suddenly gets
throttled to 400 Mhz (see below)

HW Vendor replies with "Debian ist not on the supported OS" list, so we
are currently fighting on our own.

Does someone else experince the same/similar issue? It seems to my as
some kind of thermal throttling, but kernel does not log thottling
events. Maybe some Debian-specific kernel setting, that influences CPU
throttling..


Are you sure it is not due to a "power save" feature for a system under
low load?

Good point, but no, because the system is under heavy load all the time,
not idling.
After throttling down to 400 mhz, system also stays at this speed. Only
system reboot mitigates the issue.


What CPU driver and Governor currently in use?
https://www.kernel.org/doc/html/latest/admin-guide/pm/working-state.html

#CPU driver:

root@node3:~# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver
acpi-cpufreq

#Governor:
root@node3:~# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
schedutil

I did not set/change governor/driver settings, this is a stock debian
kernel.

Is the server platform runs latest BIOS and firmware?
Things I'd try first if I was in your place.
I always flash latest firmware available as a pre-sale procedure, or 
during server installation.


I've also found this bug report¹ . Could be the same issue with scaling 
driver, which was fixed in kernel 5.11.

Debian stable runs version 5.10.84, so test the system with newer kernel.


¹ https://bugzilla.kernel.org/show_bug.cgi?id=211305

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀https://www.debian.org
⠈⠳⣄