Re: Using bind for a local caching name server, is this configuration correct?
Tim: >> It wouldn't be impossible for a modem/router to intercept DNS >> queries and put them through their own server. Stan: > I suspect that is what the router is doing. Or the ISP upstream is > monitoring traffic, and blocking inbound port 53. An option is for you to find an alternative public DNS server that listens on more than just port 53, and find one that you're not blocked accessing. Your own DNS server could be configured to forward external queries on that port, or you could use iptables rules. e.g. OpenDNS will also respond to port 5353 While service providers may think they're doing something of benefit by buggering with your network traffic, I've never found that to be the case. Censorship can obliterate wide areas of the net, not just one small problem. Trying to provide helpful alternatives instead of just failing with the correct error message causes its own set of problems. And many of these things are just so damn slow it's not funny. ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: Using bind for a local caching name server, is this configuration correct?
Stan: > $ dig example.com @1.1.1.1 > > ; <<>> DiG 9.11.8-RedHat-9.11.8-1.fc31 <<>> example.com @1.1.1.1 > ;; global options: +cmd > ;; connection timed out; no servers could be reached Tim: >> If that kind of thing fails, then you won't be able to run your own >> DNS server, there's something blocking necessary network traffic. > That explains the bind / named failure I'm seeing. You could test whether its your service provider or your router blocking that traffic. Set the router to use 8.8.8.8 as its DNS server, and see if it can still resolve names. Use the dig command with the @ portion set to the router's IP. Try and resolve a name it won't have an already cached answer for. If the router is the problem, look for its firewall settings. >> That's an odd hostname. [localhost.Home.localhost.Home] > I wonder if my thrashing around on this issue did something that is > somehow causing the change. I don't recall seeing it before. > > Jul 03 08:18:22 localhost.Home systemd-hostnamed[858]: Changed host name to > 'localhost.Home' > Jul 03 08:18:22 localhost.localdomain systemd[1]: Started Hostname Service. > Jul 03 08:18:21 localhost.localdomain NetworkManager[20472]: > [1562167101.6542] policy: set-hostname: set hostname to 'localhost.Home' > (from address lookup) > > Definitely something I did. I should probably figure out what I did > and reverse it in case there are things that expect localdomain > instead of Home. On Linux, and elsewhere, most things expect localhost and 127.0.0.1 to resolve into each other in both directions. And they are the equivalent of the computer seeing itself in the mirror, and saying "me". The computer should resolve that without resorting to a DNS server, since it's in /etc/hosts (which is consulted first). But most DNS servers have an entry for localhost, for completeness's sake. The .localdomain suffix is a Linux oddity, and I know of nothing that needs it, nowadays. In the past, I'd noticed something like sendmail demanded at least one dot in the hostname. The .home suffix is often used by routers which provide DHCP, and sometimes an integrated DNS server (where it can resolve its doled out hostnames). It was/is quite common that routers will assign hostnames with DHCP, but provide no way for the names to be resolved. Other hostnames shouldn't horse around with localhost. In other words, set a unique hostname for the computers network IP, that doesn't use the word. Give services the hostname that other devices will find them at, not localhost (that name only applies to itself, all other devices will also think of themselves as localhost). e.g. /etc/hosts #IP namealternative names (aliases) 127.0.0.1localhost localhost.localdomain 192.168.0.1 backup backup.lan backup.home backup.localdomain 192.168.0.2 filestore.home filestore Just giving multiple examples of how can be used. The first name after the IP is the one given if something checks for the hostname of an IP. The other names will also be associated with the IP. With a DNS server, the hosts file only needs to carry the 127.0.0.1 line, all other names and IPs would be resolved by the DNS server. >> Link-local addresses are not resolved by traditional DNS servers, on >> port 53. It uses multicast DNS (mdns) on a different port, such as >> 5353. It runs independently, and is not co-operative with >> traditional DNS. > That doesn't sound like my system, since I get assigned a dhcp > address by the router. Without testing, I couldn't say what scheme/schemes your router supports. So many things use it now, such as home network printers. If it's enabled in your computer, it can use all of these different schemes. And the first one to answer wins. So slow DHCP servers will be a problem. ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: info files index generation on fedora 30 - how does the 'dir' file get generated?
Hi, Ronaldo Mercado wrote: > I think the packages in old versions of fedora had a %postinstall script > that called /sbin/install-info. Those scripts seem to be gone. > > How does the /usr/share/info/dir generation work on fedora 30? Individual package scriptlets have been replaced with a file trigger in the info package. It runs whenever another package places files under the info dir. https://src.fedoraproject.org/rpms/texinfo/blob/f30/f/texinfo.spec#_100 Hope that helps, -- Todd ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: info files index generation on fedora 30 - how does the 'dir' file get generated?
On Fri, 05 Jul 2019 00:31:50 +0100 Ronaldo Mercado wrote: > Yesterday I noticed that the info files from /usr/share/info/*.info.gz > on my fedora 30 pc were missing. All the info files are there and in the directory on my f30 system, though I couldn't say where they came from. I haven't installed anything that uses info outside of the normal f30 rpms. ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
info files index generation on fedora 30 - how does the 'dir' file get generated?
Hi, Yesterday I noticed that the info files from /usr/share/info/*.info.gz on my fedora 30 pc were missing. This was because emacs 'info' command reported "emacs can't find the info directory node" A while the /usr/share/info/dir file had been generated. What generated the info directory file /usr/share/info/dir? I think the packages in old versions of fedora had a %postinstall script that called /sbin/install-info. Those scripts seem to be gone. How does the /usr/share/info/dir generation work on fedora 30? Thanks ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: sound device initialization?
Have you tried using pavucontrol? I have an issue with sound on my laptop with the MATE desktop and an external monitor. Often when I login to the laptop I'll have no sound and the mate_volume_applet doesn't change the volume. Using pavucontrol I determined that the problem was the monitor plugged into the HDMI port of my laptop. pavucontrol shows 2 entries under the configuration tab for built in audio, one for the HDMI port and one for the actual built in audio. As soon as I disable the HDMI entry the mate_volume_applet works and I have sound. That is until for some reason the HDMI entry gets turned on again and I have to disable it again. Paolo On 7/4/19 8:53 AM, Tom Horsley wrote: Lately when I install updates and reboot, I'll randomly have no sound when I log back in. I'm supposed to get sound output on my HDMI interface (being run by the nvidia binary driver from rpmfusion). If I run "gnome-control-center sound", it will show as currently using a dummy device, and if I try to switch to HDMI output (which is listed as a choice), it won't allow me to test the audio out. Rebooting and hoping I'll have sound when I come back has been the only way I've found to fix this so far. Anyone know if there is some way to re-initialize audio devices less drastic than rebooting? ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: Kernel entries are not updated in grub menu
I've had trouble too is a new Fedora install with grub on the partition (not MBR), I even eventually re-installed because I tried to revert to grubby-deprecated and then couldn't get the system to boot to graphical.target because of the entries needed for nvidia drivers, even after reverting to nouveau. So now, although there hasn't been no new kernels yet, my mantra will be don't touch what's working. I'll just do the manual grub2-mkconfig... As for the new bls, there are some meagre instructions at https://fedoramagazine.org/setting-kernel-command-line-arguments-with-fedora-30/ Check 'grubby' manual, maybe. ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: UDEV RULES
You should be using the UUID of the partition to detect and mount your target. Detecting the "whole" drive is immaterial, since you want udev to mount a partition, not a drive. I would try to get one partition working first, then proceed to the others. With udev testing, I've always had far more luck testing individual tasks, one by one. Throw too much stuff at it, and you can't tell what's not working: udev, your script, or you. On Thu, Jul 4, 2019 at 3:26 AM Angelo Moreschini wrote: > HI Seth, > > I'm sorry to still bother you ..., but your help could save me a lot of > useless tests (since there are many parameters to manage and knowing what I > really have to do could, this trigger me the use of useless ones). > > > I think that I understood the crux of the problem ... (the UDEV theory …) > but I still have a doubt that comes from the division of my HD into > partitions (to how this particularity can influence my particular problem). > > *I would like exactly to know if UDEV reveals the existence of the entire > HD USB, or if instead it detects only the individual existence of the three > partitions that are mounted there ..* > > > Because I read this.., I think that I could be able to check the real > “work” of UDEV on my computer : > > *“Whenever you connect to the system, the kernel detects and initializes > it, and directory with the device name is created under / sys / directory > which stores the device attributes.”.* > > This could be a great help, but I couldn’t identify “this: > sub-directory…., in the /sys directory, they are there many directories > with many ramifications , how I can to identify the one in which they are > stored the attributes of the device whose presence is detected by UDEV? > > > Knowing if UDEV only reveals the existence of the (three) partitions and > not the existence of the entire HD also has reflections on the script > script that uses rsynk to carry out the bak-up. > > This is because (probably) I should write three scripts that use the > utility rsync (one script for each partition) and not a single script > that acts on the three different partitions. > > > Thank you > > regards > > > Angelo > > > On Wed, Jul 3, 2019 at 12:40 AM Fred Smith > wrote: > >> On Wed, Jul 03, 2019 at 07:16:17AM +1200, Seth Kenlon wrote: >> > Using the UUID is correct. >> > >> > If the script isn't being executed by udev but you can confirm that >> > udev is correctl mounting your target partition, and the shell script >> > functions as expected when you run it manually, then I would create a >> > simple timestamp shell script just to confirm that you can get a shell >> > script to do *something* due to a udev event. The more complex the >> > shell script, the more variables you have to debug, so simplify your >> > shell script and work to just get a time stamp (or something) when a >> > udev event occurs. >> >> I've done that by embedding echo commands into the script that >> will put debugging statements a debug file that you can look >> at after bootup. >> >> put one statement before anything else in the script and if you >> don't get any debugoutput you'll know the script isn't being run. >> >> Fred >> > >> > On Tue, Jul 2, 2019 at 3:07 AM Angelo Moreschini >> > wrote: >> > > >> > > thank you Seth, >> > > >> > > I have no problems with the script executed at the UDEV event (if >> this is the meaning you attribute to the term ). >> > > I've already done this script that works, if run from a terminal. >> > > If by the term you mean something different, please >> tell me. >> > > >> > > >> > > The problem is that this script is not executed as a result of the >> UDEV event. >> > > - - - - - - >> > > But maybe there is still something not clear for me that I can ask >> you: >> > > The HD (USB) that I am using, is divided into three partitions (in >> order to keep the saved data separate): >> > > = - = - = - = - = - = >> > > the blkid command give this output ...: >> > > / dev / sdc1: LABEL = "PRTZm_sys-admin" UUID = "1A5CEE7149E4BA81" >> TYPE = "ntfs" PARTUUID = "0007fd62-01" >> > > / dev / sdc2: LABEL = "PRTZm_data-common" UUID = "7FB1E3D91206DE1C" >> TYPE = "ntfs" PARTUUID = "0007fd62-02" >> > > / dev / sdc3: LABEL = "PRTZm_programming" UUID = "0A69FDE9688969D9" >> TYPE = "ntfs" PARTUUID = "0007fd62-03" >> > > = - = - = - = - = - = >> > > so far, till now, I tried to reveal the montage of the HD without >> taking into the account of partitions into which the USB HD is divided .. >> > > But since the blkid command does not reveal its own UUID I am now >> thinking that this could be the problem. >> > > >> > > IN CONCLUSION, since the USB is divided into partitions, ... will the >> UDEV event reveal the presence of the entire HD (as the sole unit), or will >> it instead reveal the presence of the three individual different partitions? >> > > >> > > Still I would ask : this is a good syntax ? >> ENV{ID_FS_UUID}=="1A5CEE7149E4BA81" >> > > and still also, :-) >> >
sound device initialization?
Lately when I install updates and reboot, I'll randomly have no sound when I log back in. I'm supposed to get sound output on my HDMI interface (being run by the nvidia binary driver from rpmfusion). If I run "gnome-control-center sound", it will show as currently using a dummy device, and if I try to switch to HDMI output (which is listed as a choice), it won't allow me to test the audio out. Rebooting and hoping I'll have sound when I come back has been the only way I've found to fix this so far. Anyone know if there is some way to re-initialize audio devices less drastic than rebooting? ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: external USB drives
On 7/4/19 4:33 AM, wwp wrote: Hello, On Wed, 3 Jul 2019 15:42:31 -0700 ToddAndMargo via users wrote: On 7/3/19 1:50 PM, Paolo Galtieri wrote: Folks, I recently purchased a 4TB external USB drive. I have purchased USB > drives in the past and have run dosfslabel to set the drive label. When > I tried this on F29 I get the following: dosfslabel /dev/sdc2 NEW_LABEL Currently, only 1 or 2 FATs are supported, not 0. I have 2 8TB drives which I labeled some time ago without a problem, but > now I get the same error when I try to print out the drive label dosfslabel /dev/sdd2 Currently, only 1 or 2 FATs are supported, not 0. The new drive shows up on the desktop as: "Seagate Expansion Drive" I tried the same thing on F30 with the same result. Can someone tell me what's going on? Paolo Hi Paolo, It is probably formatted in NTFS. For the sake of the archives, I've just bought several ones too (the 5TB model), they are seen by gparted as exFat. Reformatted to ext4, tune2fs'ed and put in production. gparted is our friend. ! ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Re: external USB drives
Hello, On Wed, 3 Jul 2019 15:42:31 -0700 ToddAndMargo via users wrote: > On 7/3/19 1:50 PM, Paolo Galtieri wrote: > > Folks, > > I recently purchased a 4TB external USB drive. I have purchased USB > > > drives in the past and have run dosfslabel to set the drive label. When > > > I tried this on F29 I get the following: > > > > dosfslabel /dev/sdc2 NEW_LABEL > > > > Currently, only 1 or 2 FATs are supported, not 0. > > > > I have 2 8TB drives which I labeled some time ago without a problem, but > > > now I get the same error when I try to print out the drive label > > > > dosfslabel /dev/sdd2 > > > > Currently, only 1 or 2 FATs are supported, not 0. > > > > The new drive shows up on the desktop as: > > > > "Seagate Expansion Drive" > > > > I tried the same thing on F30 with the same result. > > > > Can someone tell me what's going on? > > > > Paolo > > Hi Paolo, > > It is probably formatted in NTFS. For the sake of the archives, I've just bought several ones too (the 5TB model), they are seen by gparted as exFat. Reformatted to ext4, tune2fs'ed and put in production. Regards, -- wwp pgpFakFNwyJJM.pgp Description: OpenPGP digital signature ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org