Re: How do I set the TCP Initial Window size to 10-50KB? (i.e. how configure TCP Slow-Start/Anti-Congestion to burst 10-50KB w.o. interference)

On 2015-03-29 16:59, Stuart Henderson wrote: On 2015/03/29 14:45, Tinker wrote: On 2015-03-29 13:50, Stuart Henderson wrote: On 2015-03-29, Tinker ti...@openmailbox.org wrote: How do I make TCP stack's initial window 10-50KB, so that I can burst that amount of data without any interference from

Kernel enforces MTU if set as initial [congestion] window size and leaves no config option, is that really good idea? How know that the automtu setting is off for an IPv4 NIC? Please correct me, tha

On 2015-03-29 13:50, Stuart Henderson wrote: On 2015-03-29, Tinker ti...@openmailbox.org wrote: How do I make TCP stack's initial window 10-50KB, so that I can burst that amount of data without any interference from the TCP slow-start/RFC 3390 anti-congestion whatsoever? OpenBSD already

Re: How do I set the TCP Initial Window size to 10-50KB? (i.e. how configure TCP Slow-Start/Anti-Congestion to burst 10-50KB w.o. interference)

, so that the send at Y would be speed-capped/blocked? Thanks, Tinker

How do I set the TCP Initial Window size to 10-50KB? (i.e. how configure TCP Slow-Start/Anti-Congestion to burst 10-50KB w.o. interference)

that would be a key.) Thanks!! Tinker

Re: How do I set the TCP Initial Window size to 10-50KB? (i.e. how configure TCP Slow-Start/Anti-Congestion to burst 10-50KB w.o. interference)

On 2015-03-29 13:50, Stuart Henderson wrote: On 2015-03-29, Tinker ti...@openmailbox.org wrote: How do I make TCP stack's initial window 10-50KB, so that I can burst that amount of data without any interference from the TCP slow-start/RFC 3390 anti-congestion whatsoever? OpenBSD already

Re: Will mmap and the read buffer cache be unified, anyone working with it?

Okay there's msync(). So this really doesn't look like a problem but let me think. Any thoughts on this one welcome. On 2015-05-22 12:57, Tinker wrote: Hi, Will mmap and the read buffer cache be unified, anyone working with it? Some programs disable features on OBSD for this reason so would

Re: Will mmap and the read buffer cache be unified, anyone working with it?

On 2015-05-23 16:42, Stuart Henderson wrote: On 2015-05-22, Tinker ti...@openmailbox.org wrote: Okay there's msync(). So this really doesn't look like a problem but let me think. It is a problem because it's not needed on the majority of other OS where most upstream testing is done, so even

Will mmap and the read buffer cache be unified, anyone working with it?

Hi, Will mmap and the read buffer cache be unified, anyone working with it? Some programs disable features on OBSD for this reason so would be nice! (I admit though that a program combining mmap() and read() on the same file sounds like a slightly quirky design choice to me.) Thanks! Tinker

Re: Will mmap and the read buffer cache be unified, anyone working with it?

On 2015-11-12 01:30, Theo de Raadt wrote: I find this conversation puzzling, since even back in BSD 4.3, read() was actually implemented by memory mapping the underlying file. That is an wildly incorrect description of the 4.3 BSD buffer cache, furthermore 4.3 lacked a working mmap() to hit

Re: Will mmap and the read buffer cache be unified, anyone working with it?

... LMDB's write performance is pretty mediocre, by design - we emphasized durability/reliability over performance here. But in practice, it is always faster than e.g. BerkeleyDB, which supports multiple concurrent writers. With multiple writer concurrency, we found that BDB spends much of its

Re: Will mmap and the read buffer cache be unified, anyone working with it?

, and that the NFS filesystem drivers had been what stopped it then, by some reason they were difficult. Perhaps that could be worked around afterall? On 2015-05-22 15:27, Tinker wrote: Hi, Will mmap and the read buffer cache be unified, anyone working with it? Some programs disable features on OBSD

Re: Will mmap and the read buffer cache be unified, anyone working with it?

On 2015-11-10 14:10, Tinker wrote: ... A "safe" approach to file access would be to read data using mmap() but write data using fwrite() only. Mmap does have a read-only mode. This does NOT work in OpenBSD currently though because of the absence of unified caching. The nice thing abo

Re: Will mmap and the read buffer cache be unified, anyone working with it?

using mmap() but write data using fwrite() only. Mmap does have a read-only mode. This does NOT work in OpenBSD currently though because of the absence of unified caching. (Sorry for not providing a diff for this.) On 2015-11-10 13:36, Tinker wrote: Really, unifying the buffer cache woul

Re: Will mmap and the read buffer cache be unified, anyone working with it?

On 2015-11-12 01:42, Theo de Raadt wrote: On 2015-11-12 01:30, Theo de Raadt wrote: >> I find this conversation puzzling, since even back in BSD 4.3, read() >> was >> actually implemented by memory mapping the underlying file. > > That is an wildly incorrect description of the 4.3 BSD buffer

Re: Any news on Merkle tree-hash-based whole-disk checksums (=ZFS-style checksums) in softraid?

roblem of lots of reads that you mention in your email will be completely resolved - if your code is correct, then the reading overhead from your RAID1C should be almost nonexistent. Thanks, Tinker On 2015-12-02 05:15, Karel Gardas wrote: Tinker, what you basically try to describe as Fletcher i

Re: Any news on Fletcher checksums (=ZFS-style checksums) in softraid?

Wait, so you say the input for your CRC32 checksum is "metadata>". So every sector/block in your model contains a CRC32 checksum of that, and on every fread() you check that. Does the SR metadata contain the sector index number, so that if the sector index number would have changed

Re: Any news on Fletcher checksums (=ZFS-style checksums) in softraid?

hole point with it, that you're in a place where there always are backups and you just want to maximize the read correctness guarantees. For anything important I'd easily prefer to use that. On 2015-12-02 03:40, Tinker wrote: Just to illustrate the case. This is just how I got that it works, please pardon t

Re: Any news on Merkle tree-hash-based whole-disk checksums (=ZFS-style checksums) in softraid?

On 2015-12-02 07:14, Raul Miller wrote: On Tue, Dec 1, 2015 at 5:21 PM, Tinker <ti...@openmailbox.org> wrote: So your current solution is *NOT* data-safe toward "mis-write":s and other write errors that go unnoticed at write time. While I agree that the probability that th

Re: Any news on Merkle tree-hash-based whole-disk checksums (=ZFS-style checksums) in softraid?

Karel, the most important thing at the bottom of the email :) On 2015-12-02 19:10, Karel Gardas wrote: To answer your question: In that case, as soon as that invalid data would actually be read from disk, it would be caught by the checksums that are guaranteed to be kept in RAM, so that is,

+1 vote for including the keydisk-on-same-hdd patch :) Re: Unable to sufficiently clean up softraid metadata (Re: crypto softraid and keydisk on same harddrive)

On 2015-12-02 08:26, Patrik Lundin wrote: Hello, I have a custom installer script which automatically creates RAID devices and assembles an sd1 CRYPTO device before the ordinary installer continues (making the installer use sd1 for the rest of the installation). This works well, other than

Any news on Fletcher checksums (=ZFS-style checksums) in softraid?

://en.wikipedia.org/wiki/Fletcher%27s_checksum Thanks! Tinker

Re: Any news on Merkle tree-hash-based whole-disk checksums (=ZFS-style checksums) in softraid?

On 2015-12-02 18:10, Karel Gardas wrote: On Wed, Dec 2, 2015 at 10:39 AM, Tinker <ti...@openmailbox.org> wrote: On 2015-12-02 17:31, Karel Gardas wrote: I think RAID1C is capable of detecting mis-directed write. I wrote about it in some of my previous email. Hi Karel, I'll fol

Re: Any news on Merkle tree-hash-based whole-disk checksums (=ZFS-style checksums) in softraid?

! Tinker

Any news on Fletcher checksums (=ZFS-style checksums) in softraid? (+better phrasing)

prompted to use backups, instead of going into processing broken information, which could have unlimitedly bad consequences (crash programs, compromise information, etc.). So it's really like a night and day difference. https://en.wikipedia.org/wiki/Fletcher%27s_checksum Thanks! Tinker

Re: Any news on Merkle tree-hash-based whole-disk checksums (=ZFS-style checksums) in softraid?

t want a catch-all data safety mechanism that safeguards against every type of disk breakdown, that's all. What do you say about this way of reasoning? Thanks, Tinker https://en.wikipedia.org/wiki/Btrfs#Checksum_tree_and_scrubbing https://en.wikipedia.org/wiki/HAMMER On 2015-12-02 10:17, Raul Mi

Re: Any news on Fletcher checksums (=ZFS-style checksums) in softraid? (+better phrasing)

On 2015-12-02 02:27, Chris Cappuccio wrote: Tinker [ti...@openmailbox.org] wrote: Hi! I heard someone was working with implementing Fletcher checksums in softraid. Do you know any updates on this? Karel Gardas was working on an implementation of RAID1C for softraid Last I remember

Re: Any news on Fletcher checksums (=ZFS-style checksums) in softraid?

. the Fletcher thing would catch *any* inconsistency anywhere on the disk. Maybe it could be argued that it's "too picky" for some less data-safe environments, but, in a place where you have good backups and you value 100.0% fread() correctness, it's awesome!! Looking forward to

Re: Any news on Fletcher checksums (=ZFS-style checksums) in softraid?

and the 1200-1300MB checksum checks would fail. Reading other parts of the disk would work though. On 2015-12-02 03:31, Tinker wrote: Hi Karel, Glad to talk to you. Why the extra IO expense? About the Fletcher vs not Fletcher thing, can you please explain to me what happens in a setup where

What are the first 2 columns in "top"'s memory report for really? Man page doesn't say!

ther stuff. Is it that the kernel with all of its work takes 414MB?? If so that's weird because really the machine doesn't do a lot. The "man" page doesn't say any of this. I think it's good to know this, for diagnostic purposes. Anyone knows what the second column is for? Thanks, Tinker

Re: What are the first 2 columns in "top"'s memory report for really? Man page doesn't say!

On 2015-12-22 03:32, Otto Moerbeek wrote: On Tue, Dec 22, 2015 at 02:31:58AM +0800, Tinker wrote: Hi, on my 4GB RAM machine, top says "Memory: Real: 1293M/3786M act/tot Free: 158M Cache: 2079M Swap: 449M/10G" All the five first columns vary over time, in particular the first fo

Re: Boot loader uses INT 13h [WAS BIOS call fallback]

On 2015-12-23 10:04, Dragos Ruiu wrote: Ok let me short circuit this meta discussion by saying that AFAIK now that the new Intel Skylake chips fixed many virtualization bugs Curious, where can I read about this, URL? and it's possible to efficiently nest VMs there might not be a way to

Re: Boot loader uses INT 13h [WAS BIOS call fallback]

On 2015-12-23 18:14, Dragos Ruiu wrote: Sure you could spend the rest of your life checking all the firmware and trying to design separate specialized tools for the myriad of devices in a modern PC - and there is a lot more than your simple list, see the presentation Mickey Shkatov and Jesse

Re: "bioctl -P" is to change passphrase without wiping the encrypted partition's contents. How do you generate a new keydisk without wiping the same?

previous password/keydisk. Would they be able to extract any part of the disk information then, if not why? On 2015-11-20 21:58, Tinker wrote: "bioctl -P" is to change passphrase without wiping the encrypted partition's contents. How do you generate a new keydisk without wipin

"bioctl -P" is to change passphrase without wiping the encrypted partition's contents. How do you generate a new keydisk without wiping the same?

d. How do you generate a new one without needing to wipe /dev/sd0a . I.e. exactly the same as "-P" but for the keydisk usecase. (Of course the old keydisk/password is needed at replacement time.) Thanks, Tinker

Re: "bioctl -P" is to change passphrase without wiping the encrypted partition's contents. How do you generate a new keydisk without wiping thesame?

the space to backup 100% of the disk as to start over. On 2015-11-21 02:45, Ted Unangst wrote: Tinker wrote: Ah, and maybe equally importantly, what are the security ramifications of changing password/keydisk vs. wiping and installing from scratch with a new password/keydisk? The master key

Re: "bioctl -P" is to change passphrase without wiping the encrypted partition's contents. How do you generate a new keydisk without wipingthesame?

the space to backup 100% of the disk as to start over. On 2015-11-21 03:03, Ted Unangst wrote: Tinker wrote: Aha. *Is* the keydisk the master key, and hence can't be changed? The keydisk is the mask for the master key. It can (in theory) be changed like changing a password. Really, the key

Any update on LMDB support for OpenBSD?

Hi Howard, Any plan on making LMDB work on OpenBSD? OpenBSD should deliver LMDB in the MDB_WRITEMAP mode. (http://lmdb.tech/doc/group__mdb.html) (At least some effort has been done e.g. https://gitlab.com/mdb/sqlightning/issues/6 ) Thanks, Tinker

Re: Any update on LMDB support for OpenBSD?

Howard, find three followup Q:s below (3-5), thanks On 2016-06-13 05:07, Howard Chu wrote: Tinker wrote: Hi Howard, Any plan on making LMDB work on OpenBSD? Not sure what action you're expecting from us. Theo made it quite clear that OpenBSD does not and will not have a unified buffer

Can read-only mmap() and fwrite() be combined.. via msync or something? In OpenBSD now. (repetition of Q sorry)

? Thanks! Tinker

Will Softraid RAID1 read from the fastest mirror/-s / supports user-specified device read priority order, nowadays? Takes broken disk out of use?

a read/write failure (or excessive time consumption for a single operation, say 15 seconds), will Softraid RAID1 learn to take the broken disk out of use? Thanks, Tinker

How extensive OpenBSD's write caching (for softdep or async-mounted UFS, as long as I never fsync() )?

ta that's been written but still only is in the OS RAM CACHE) etc. might be so slow that a program would need to implement its own write cache for supporting even small spikes in write activity. Sorry for the fuss but neither "man" nor googling taught me anything. Thanks!! Tinker

Buffer cache made to use >32bit mem addresses (i.e. >~3GB support for the buffer cache) nowadays or planned soon?

to check, has this cap been removed, or is there any plan to remove it next months from now? Thanks, Tinker (Sorry for the spam, hope it serves some constructive purpose.) [1]: * http://unix.stackexchange.com/questions/61459/does-sysctl-kern-bufcachepercent-not-work-in-openbsd-5-2-above-1-7gb

Re: Buffer cache made to use >32bit mem addresses (i.e. >~3GB support for the buffer cache) nowadays or planned soon?

re have it right? Thank you everyone for your excellent work with OpenBSD! Best regards, Tinker [1] David Mazieres linked to two docs in there, those are only on archive.org now: https://web.archive.org/web/20150814051509/http://www.intel.com/content/dam/www/public/us/en/documents/product-s

Re: Buffer cache made to use >32bit mem addresses (i.e. >~3GB support for the buffer cache) nowadays or planned soon?

upport? Are there plans today to implement this and then get the >32bit/>~3GB support going? (I don't understand exactly how these things fit together, this is why I asked tentatively.) Thanks, Tinker On 2016-02-14 02:03, Karel Gardas wrote: I'm afraid you read too quickly and w/o attention to de

Re: Can I accelerate my magnet HDD using a SSD in any way?? E.g. softraid patch/ARC, dedicated hardware e.g. Intel RCS25ZB040LX="Nytro MegaRAID", anything

Patrick, To sum up, neat to see that (from what we can see without having tested it,) there is (even inexpensive) hardware for this on the market, neat! My last question related to this would be, what if drives start breaking down (storage or CacheCade drives), would the OpenBSD system

Re: Can I accelerate my magnet HDD using a SSD in any way?? E.g. softraid patch/ARC, dedicated hardware e.g. Intel RCS25ZB040LX="Nytro MegaRAID", anything

k & SSD with super cap / fault tolerant on disk cache can seed up I/O significantly.. Can you give a practical example of this? ) Thanks, Tinker

Re: Can I accelerate my magnet HDD using a SSD in any way?? E.g. softraid patch/ARC, dedicated hardware e.g. Intel RCS25ZB040LX="Nytro MegaRAID", anything

On 2016-02-01 16:33, Janne Johansson wrote: 2016-01-31 9:16 GMT+01:00 Tinker <ti...@openmailbox.org>: This could be made in software with benefit, as a Softraid patch. So the frequently accessed stuff ends up cached on the SSD for faster read speed. There is some hardware solutio

Re: Karel, some followup Q:s on your RAID1C patch

Janne, ZFS includes the sector number in the checksum input because that provides a safeguard against any form of disk breakdown. Most filesystems except for ZFS are susceptible to this kind of error because either it's not checksummed, or the checksums are locally stored so data

Can I accelerate my magnet HDD using a SSD in any way?? E.g. softraid patch/ARC, dedicated hardware e.g. Intel RCS25ZB040LX="Nytro MegaRAID", anything

This could be made in software with benefit, as a Softraid patch. So the frequently accessed stuff ends up cached on the SSD for faster read speed. ZFS on FreeBSD etc. does it in its "ARC"/"ARC2L" feature? There is some hardware solution, e.g. Intel made the

Re: Can I accelerate my magnet HDD using a SSD in any way?? E.g. softraid patch/ARC, dedicated hardware e.g. Intel RCS25ZB040LX="Nytro MegaRAID", anything

for making this happen, if so how big a donation, of what value would USD 1,000 be in proportion? Not sure right now but next 12 months I could consider this, if you have any ideas about who feel free to respond or PM or anything you like. On 2016-01-31 16:16, Tinker wrote: This could be made

Re: Can I accelerate my magnet HDD using a SSD in any way?? E.g. softraid patch/ARC, dedicated hardware e.g. Intel RCS25ZB040LX="Nytro MegaRAID", anything

Ah, I now understand that this problem is mindbogglingly complex because of tons and tons of work needed to make it work, including the storage format on the SSD cache, and tools for "fsck":ing it etc etc. In a way that maybe answered my question, thanks! On 2016-01-31 19:00, Ti

Re: Karel, some followup Q:s on your RAID1C patch

On 2016-02-02 01:23, Karel Gardas wrote: Tinker, adding block number to sector data and calculate chksum over this should be relatively easy. I'll ping you once I send another WIP patch to tech@ so you will get a chance to hack on it... Besides this, I've observed you are huge fan of ZFS

Just curious, what are approx plans for kernel-multicore-ness next 1-2 years? (net/file IO) How much is it a relevant prio?

First, Thank you for this fantastically awesome OS. Just curious, what are approx plans for kernel-multicore-ness next 1-2 years? (Network IO, file/disk IO) Are there any relevant cases today where the singlecore architecture actually is limiting? Say on a 20-core machine doing random file

What's good safe PPC/MIPS/SPARC networking hardware with open firmware that works well with OpenBSD? Many ethernet plugs and 1U rack mount = bonus.

Hi! What's good PPC/MIPS/SPARC networking hardware with open firmware that works well with OpenBSD? Minimalistic for minimizing attack surface. (Many ethernet ports would be a bonus. 1U rack mount would be a bonus. ECC would be a bonus.) Thanks! Tinker

Re: What's good safe PPC/MIPS/SPARC networking hardware with open firmware that works well with OpenBSD? Many ethernet plugs and 1U rack mount = bonus.

(es). fw?/app server?/storage?/nas? etc. Network infrastructure, so, among those categories it would be FW. What hardware is advisable here? On Sat, Feb 27, 2016 at 11:10 PM, Tinker <ti...@openmailbox.org> wrote: Hi! What's good PPC/MIPS/SPARC networking hardware with open fi

Re: Just curious, what are approx plans for kernel-multicore-ness next 1-2 years? (net/file IO) How much is it a relevant prio?

On 2016-02-25 19:05, Jiri B wrote: On Thu, Feb 25, 2016 at 06:14:40PM +0700, Tinker wrote: Any guess about when might it actually become usable, 6.1, 6.2, this year, next? I would be very curious to see if-what limits there are on file IO activity when you do it full-on completely RW on 10

Re: Just curious, what are approx plans for kernel-multicore-ness next 1-2 years? (net/file IO) How much is it a relevant prio?

On 2016-02-25 06:04, Chris Cappuccio wrote: Tinker [ti...@openmailbox.org] wrote: First, Thank you for this fantastically awesome OS. Just curious, what are approx plans for kernel-multicore-ness next 1-2 years? (Network IO, file/disk IO) Network SMP (and other subsystems required

Softraids not sanwhichable? Re: Can I accelerate [...]

Dear Karel, Are you saying that softraids not can be sandwhiched as of today - If they can't, about what kind of complexity needs to be solved? Thanks, Tinker On 2016-01-31 19:06, Karel Gardas wrote: [..] If you are interested in softraid, then perhaps some funding of fixes which would allow

Re: How extensive OpenBSD's write caching (for softdep or async-mounted UFS, as long as I never fsync() )?

respectively, which is the harddrive's write speed - the buffer cache of course would have absorbed this in 0 seconds.) So, both runs showed that OpenBSD *not* does any write caching to talk about, at all. This means if a program wants write caching, it needs to implement it itself. Good to know. Tink

Re: Buffer cache made to use >32bit mem addresses (i.e. >~3GB support for the buffer cache) nowadays or planned soon?

On 2016-02-14 03:39, Stuart Henderson wrote: On 2016-02-13, Tinker <ti...@openmailbox.org> wrote: Hi, Some quite deep reading [1] taught me that at least quite recently, there was a ~3GB cap on the buffer cache, independent of architecture and system RAM size. Reading the source h

Re: How extensive OpenBSD's write caching (for softdep or async-mounted UFS, as long as I never fsync() )?

On 2016-02-14 19:20, Otto Moerbeek wrote: On Sun, Feb 14, 2016 at 06:56:16AM -0500, Donald Allen wrote: On Sun, Feb 14, 2016 at 1:43 AM, Tinker <ti...@openmailbox.org> wrote: > Did two tests, one with async and one with softdep, on amd64, 5.9-CURRENT, > UFS. > > (Chec

Re: Buffer cache made to use >32bit mem addresses (i.e. >~3GB support for the buffer cache) nowadays or planned soon?

On 2016-02-15 10:15, Constantine A. Murenin wrote: .. I think it got reverted by: .. but I'm not an expert so would wait on confirmation by Bob Beck. Yes, I think you are correct, and it was indeed reverted. .. But it looks like the functions that were introduced in the above commit are

Re: Will Softraid RAID1 read from the fastest mirror/-s / supports user-specified device read priority order, nowadays? Takes broken disk out of use?

Dear Nick, On 2016-02-15 05:29, Nick Holland wrote: On 02/13/16 11:49, Tinker wrote: Hi, 1) http://www.openbsd.org/papers/asiabsdcon2010_softraid/softraid.pdf page 3 "2.2 RAID 1" says that it reads "on a round-robin basis from all active chunks", i.e. read operatio

Re: Will Softraid RAID1 read from the fastest mirror/-s / supports user-specified device read priority order, nowadays? Takes broken disk out of use?

Constantine, Just basically followup to say that I agree with you. On 2016-02-15 17:41, Constantine A. Murenin wrote: On 13 February 2016 at 08:50, Tinker <ti...@openmailbox.org> wrote: Hi, 1) http://www.openbsd.org/papers/asiabsdcon2010_softraid/softraid.pdf page 3 "2.2 R

Storage server HW advice/feedback req for setup overall & in particular reliability/QoS of SATA, to protect from controller- or BIOS-induced system crashes? Dedicated PCI SATA HBA needed??

-adapters/ and www.avagotech.com/products/server-storage/raid-controllers/ . If so, what do you say about SATA HW choice? (Later on I'll need to learn about softraid hot spares, scrubbing, rebuild and hotswapping, though those are software questions.) Thanks! Tinker

Re: Storage server HW advice/feedback req for setup overall & in particular reliability/QoS of SATA, to protect from controller- or BIOS-induced system crashes? Dedicated PCI SATA HBA needed??

On 2016-02-17 01:01, j...@bitminer.ca wrote: .. Why do you think you need to build such a device? Why don't you buy it? (Dell PowerEdge VRTX, HP hyper converged, etc) Colocation requires rack servers, but thanks for thinking about it. Some important things: - what is the purpose of this

OpenBSD softraid can do scrub, hotspare, hotswap? How do rebuild + those 3 really done? (Absence of docs and howtos - ultimate Q!)

t get going at all, should I just reset from backup, or is there some kind of "rescue rebuild" I could attempt? k) Do I need to know anything about DUID vs. device name use within the softraid? Looking forward to your explanation, thanks!! :DD Tinker

..And: Will softraid support !=512byte sectors anytime soon? (SSD:s frequently have 4KB or 16KB)

http://www.openbsd.org/cgi-bin/man.cgi/OpenBSD-current/man8/bioctl.8?query=bioctl=i386 says "CAVEATS Only devices with 512-byte sectors are supported." Is any update upcoming? While it's 512-byte only, what does that mean for write wear amplification and access speeds on SSD:s with 4KB or

Re: OpenBSD softraid can do scrub, hotspare, hotswap? How do rebuild + those 3 really done? (Absence of docs and howtos - ultimate Q!)

On 2016-02-20 22:23, Marcus MERIGHI wrote: ti...@openmailbox.org (Tinker), 2016.02.20 (Sat) 15:29 (CET): This email is an attempt to get some knowledge on how softraid works. So many of your questions are answered if you start with bioctl(8)[1], and continue with softraid(4)[2]. Maybe bio(4

Re: OpenBSD softraid can do scrub, hotspare, hotswap? How do rebuild + those 3 really done? (Absence of docs and howtos - ultimate Q!)

...@openmailbox.org (Tinker), 2016.02.20 (Sat) 16:43 (CET): .. You appear to mean bioctl(8). Thats the only place I could find the word 'patrol'. bioctl(8) can control more than softraid(4) devices. bio(4): The following device drivers register with bio for volume management: ami

Re: OpenBSD softraid can do scrub, hotspare, hotswap? How do rebuild + those 3 really done? (Absence of docs and howtos - ultimate Q!)

On 2016-02-21 01:29, Karel Gardas wrote: scrub is IIRC not supported by any softraid yet. But there's "patrol"! "bioctl -t start mysoftraid" Rebuild by all which support redundancy. Yey! Clarified by Marcus & looking forward to his clarification Marcus recommendation to read man pages

Re: OpenBSD softraid can do scrub, hotspare, hotswap? How do rebuild + those 3 really done? (Absence of docs and howtos - ultimate Q!)

On 2016-02-21 05:05, Karel Gardas wrote: The RAID 1 discipline does not initialize the mirror upon creation. This is by design because all sectors that are read are written first. There is no point in wasting a lot of time syncing random data. I'm afraid the claim "all sectors that are read

Re: Can I accelerate my magnet HDD using a SSD in any way?? E.g. softraid patch/ARC, dedicated hardware e.g. Intel RCS25ZB040LX="Nytro MegaRAID", anything

telling it what's to load fast using "cat `find the-relevant-data/` > /dev/null" single-shot and perhaps via crontab, should deliver really well. Tinker

Re: Highest Speed Network Packet Generator?

On 2015-12-26 19:23, Stuart Henderson wrote: On 2015-12-26, Mohammad BadieZadegan wrote: Hi everybody, I need a network packet generator that generates Network Packets with the HIGHEST Speed! Before I migrate to OpenBSD I used PKTGEN on Linux to generate this with the

Re: Highest Speed Network Packet Generator?

On 2016-03-21 01:26, Jeremie Courreges-Anglas wrote: Is that netmap (http://info.iet.unipi.it/~luigi/netmap/) useful in OpenBSD? No. Stuart, what's your motivation for thinking so? - NetMap is a zero-copying high-performance ethernet frame IO API that works via select() and ioctl on an FD

Re: WAPBL?

On 2016-04-02 17:22, Karel Gardas wrote: .. so basically the situation is like with the current softdep which is also dangerous in slow-write-drive low-memory situation and yet it's in tree. Is "softdep" dangerous? :-O I thought it was a benevolent filesystem optimization, is it malevolent

Re: What are the disadvantages of soft updates?

Reading this "softdep" thread (that is https://marc.info/?l=openbsd-misc=142164001816142 ) was quite intriguing. Two followup questions: 1. Re EIO: I understand a disk write in softdep will compromise/crash the filesystem. But the error reported below was that it crashes the *OS*. It

Re: anti-ROP mechanism in libc

(From tech@) Conceptual nooby Q, what does this add beyond PIE & ASLR? On 2016-04-25 20:10, Theo de Raadt wrote: This change randomizes the order of symbols in libc.so at boot time. This is done by saving all the independent .so sub-files into an ar archive, and then relinking them into a new

Re: Will mmap and the read buffer cache be unified, anyone working with it?

A separate question about combining mmap access and file IO in the current absence of a unified buffer cache: If I have a readonly mmap and do fwrite to it, could I use fsync (or msync or any other call) right after the fwrite, as a tool to guarantee that the memory mapping interface is up to

Re: Performance of Firefox and Chromium

On 2016-04-30 19:23, Alan Corey wrote: Re: Performance of Firefox and Chromium Several seconds? Oh my. Try 20 minutes or more on some of the most bloated sites, Wait, what is the best guess now, did the recent scheduler patches that were posted here recently, remedy the speed issue

Re: amd5.9 mp softdep bufcachepercent

..And if you buy more RAM someday remember that the buffer cache is limited to 32 bits currently i.e. <4GB in total size. I trust that limit will be lifted someday. On 2016-04-17 11:50, Tinker wrote: Just set it to 90?? (And ensure you set the right thing i.e. kern.bufcachepercent righ

Re: amd5.9 mp softdep bufcachepercent

Just set it to 90?? (And ensure you set the right thing i.e. kern.bufcachepercent rightly spelled.) On 2016-04-17 11:43, Predrag Punosevac wrote: I rm -rf /usr/ports on the newly upgraded laptop running generic 5.9 amd MP kernel. /usr is mounted with the option softdep. It took ages for rm

Re: amd5.9 mp softdep bufcachepercent

Can you please test the kern.bufcachepercent values 10, 20 and 90 separately and have each of those individual tests based on 5 tests each for the respective setting? You would need to ensure that the buf cache had equal contents at each test iteration, so each test needs to be preceded by a

Re: Disk I/O performance of OpenBSD 5.9 on Xen

On 2016-07-14 07:27, Maxim Khitrov wrote: [...] No, the tests are run sequentially. Write performance is measured first (20 MB/s), then rewrite (12 MB/s), then read (37 MB/s), then seeks (95 IOPS). Okay, you are on a totally weird platform. Or, on an OK platform with a totally weird

Re: choosing OpenBSD for fileserver instead of FreeBSD + ZFS

On 2016-07-20 21:48, Karel Gardas wrote: ... Yes, similar functionality is in RAID1C patch I posted on tech@ in the past. Life is too intense now so I barely work on this to update it following Joel Sing requirements. Anyway, someday I hope to push it again to tech@ +1 for RAID1C!

Bare-metal PM953 / 850/950 PRO/EVO IO benchmark anyone? Re: Disk I/O performance of OpenBSD 5.9 on Xen

usecase - so first do: mount -A -o softdep -u AND do it with all FS:es in "async", so we get some idea of theoretical throughput: mount -A -o async -u Afterwards switch FS:es back to normal by: mount -A -u Thanks, Tinker

Re: Disk I/O performance of OpenBSD 5.9 on Xen

On 2016-07-13 22:57, Maxim Khitrov wrote: On Wed, Jul 13, 2016 at 10:53 AM, Tinker <ti...@openmailbox.org> wrote: On 2016-07-13 20:01, Maxim Khitrov wrote: We're seeing about 20 MB/s write, 35 MB/s read, and 70 IOPS What do you mean 70, you mean 70 000 IOPS? Sadly, no. It was actua

Re: How make "pkg_add" auto-choose some package version for me when same package is available in more versions?

On 2016-07-05 10:09, Tinker wrote: .. What made me ask this problem was Squid on 5.8, where there were two versions. I realize that two versions of a very similar package is very rare otherwise though, so maybe in the bigger picture the question didn't really make sense. Wait, can "%&quo

Re: How make "pkg_add" auto-choose some package version for me when same package is available in more versions?

-> Noise My original intention when I asked this question was for tools where I never ever will care about version number. For Squid I don't. I guess some compression tool, for instance, would count too - likely its behavior is exactly the same, and if it's not I would be fine with realizing

Re: How make "pkg_add" auto-choose some package version for me when same package is available in more versions?

On 2016-07-05 04:21, Nick Holland wrote: ( On 07/04/16 22:11, Tinker wrote: .. The *oldest* version would be fine too! Write a little script to fetch the directory of the package repository, apply /whatever logic you want/, then pkg_add that exact package. Done. The contention that &qu

libpthread on OpenBSD: Any particular requirement about HOW/WHEN to load it - is via shlib and dlopen() OK?

(Sorry for the spam.) This is a fork from the "Silly library path debug question" thread where I am figuring how to make auto-loading of dependencies work. That thread touched on a secondary question which is if there is any particular requirement on HOW/WHEN to load libpthread : A) Must

Re: libpthread on OpenBSD: Any particular requirement about HOW/WHEN to load it - is via shlib and dlopen() OK?

On 2016-07-05 14:42, Philip Guenther wrote: On Mon, Jul 4, 2016 at 11:00 PM, Tinker <ti...@openmailbox.org> wrote: ... That thread touched on a secondary question which is if there is any particular requirement on HOW/WHEN to load libpthread : A) Must libpthread be [dynamic object depe

Re: Silly library path debug question. Also, is /usr/local/lib/ in the default search path for library files?

ed by "ldd") or via dlopen(). BUT, A does not need to "ask for both B and C" - that is, "-lB -lC" or dlopen() both B and C, right - but *only* B, as long as B was compiled with "-lC" right?? It depends on what "depends" means. :-) "-

Re: What is OpenBSD's mechanism to ensure that *ONLY ONE* lib[e]stdc++ version is concurrently loaded in a process?

ially always lead to compilation failure however.) Anyhow, so, looking forward to your response to question (above and which is the same 1 & 2 below) and to wrap up with some conclusion. Thanks again! Tinker On 2016-07-07 17:02, Tinker wrote: .. 1) Are you saying that OpenBSD has *no

Any way to dump shared libraries loaded by a running process in OpenBSD?

ons of debug - or output the right thing but then terminate the process. Maybe some more mode than just on/off is in the code for LD_DEBUG but I don't think so. ) Thanks, Tinker

Re: What is OpenBSD's mechanism to ensure that *ONLY ONE* lib[e]stdc++ version is concurrently loaded in a process?

On 2016-07-08 20:31, Stuart Henderson wrote: On 2016-07-07, Tinker <ti...@openmailbox.org> wrote: [..] Having *all* library dependencies compiled with eg++ would ensure linking to the right runtime version We can't do that because one of those dependencies is mesa which in (T

Re: What is OpenBSD's mechanism to ensure that *ONLY ONE* lib[e]stdc++ version is concurrently loaded in a process?

Thank you very much for your feedback, I need to do more proper testing to understand this better (that is answer my personal bigger Q about how to do some C++ things and C++11 things in a really safe way, and part of that is to understand lib[e]stdc++ behavior), so time for me to do homework.

What is OpenBSD's mechanism to ensure that *ONLY ONE* lib[e]stdc++ version is concurrently loaded in a process?

and * A newer one, that is 4.8.X or 4.9.X, which a user would use for instance to compile C++11 code - and probably you don't need 4.8 and 4.9 concurrently. I also guess a lot of violence would be needed to have 4.8.X and 4.9.X installed concurrently as the packages have overlapping filenames.

Re: What is OpenBSD's mechanism to ensure that *ONLY ONE* lib[e]stdc++ version is concurrently loaded in a process?

On 2016-07-07 16:27, Stuart Henderson wrote: On 2016-07-07, Tinker <ti...@openmailbox.org> wrote: What is OpenBSD's mechanism for ensuring that only ONE lib[e]stdc++ version is loaded by an OS process concurrently There isn't one. So far we have mostly got lucky, but not always, for exa

Re: Make sure the MP kernel is installed by "cp /bsd.mp /bsd" or is there another better way?

Wait, "install.sh" was phased out and "install.sub" put in its place, URL http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib/miniroot/install.sub?rev=1.901=text/x-cvsweb-markup . The only lines I find that mention "bsd.mp" are: if [[ -f /mnt/bsd.mp ]] && ((NCPU > 1)); then

  1   2   >