#archlinux32 | Logs for 2019-11-07

[02:31:58] <buildmaster> pentium4/grub is broken (says rechenknecht): https://archlinux32.org
[02:32:58] <buildmaster> pentium4/dart is broken (says rechenknecht): https://archlinux32.org
[02:42:05] <buildmaster> any/riot is broken (says rechenknecht): https://archlinux32.org
[02:44:12] -!- yans has joined #archlinux32
[03:18:56] -!- thePiGrepper has quit [Ping timeout: 265 seconds]
[03:27:21] -!- thePiGrepper has joined #archlinux32
[04:33:17] -!- MrBIOS has quit [Read error: Connection reset by peer]
[04:33:49] -!- MrBIOS has joined #archlinux32
[04:50:45] <buildmaster> pentium4/android-tools are broken (says nlopc46): https://archlinux32.org
[05:26:51] -!- thePiGrepper has quit [Ping timeout: 240 seconds]
[05:27:39] -!- thePiGrepper has joined #archlinux32
[06:54:00] <buildmaster> i686/chromium is broken (says rechenknecht): https://archlinux32.org
[08:19:05] -!- yans has quit [Ping timeout: 265 seconds]
[08:23:49] -!- deep42thought has joined #archlinux32
[08:23:49] <buildmaster> Hi deep42thought!
[08:23:49] <buildmaster> !rq deep42thought
[08:23:51] <phrik> buildmaster: <deep42thought> computers ranked by dirt: smokers, car mechanics, bedroom, ...
[08:47:16] -!- DepositePirate has quit [Ping timeout: 260 seconds]
[08:49:29] -!- DepositePirate has joined #archlinux32
[11:36:05] -!- unrealeagle has joined #archlinux32
[11:42:18] unrealeagle is now known as aquilairreale
[11:55:54] <aquilairreale> I fixed a package (xaos) that currently failed to build. Is there a standard way to contribute the patch to the project so we may have xaos's package back in the repos? (because at this time it's missing altogether, maybe it never built successfully)
[11:58:56] <deep42thought> the standard way *would* be to open pull-requests on git.archlinux32.org
[11:58:59] <deep42thought> but that one is down
[11:59:16] <deep42thought> so just `git send-email` me the patch, and I can apply it
[12:00:22] <deep42thought> to deep42thought@archlinux32.org - in case, that was not obvious ;-)
[12:03:24] <aquilairreale> That's what I thought... btw this server outage's been pretty nasty, it's been going on for quite some time now
[12:03:32] <deep42thought> yes
[12:03:37] <deep42thought> tyzoid is not available
[12:03:51] <deep42thought> I have no clue, what happened
[12:04:16] <deep42thought> we're looking into higher-fault-tolerant alternatives atm
[12:04:30] <deep42thought> e.g.: abaumann looks into it
[12:08:00] <aquilairreale> Hmm, I understand. Wish you best of luck with that; I'll be mailing you that patch
[12:08:11] <deep42thought> thanks
[12:32:18] <aquilairreale> Hm, wait a minute, I used asp32 export to pull the original build files; it got them from archlinux's svntogit. Is that the right repo to edit and mail you the patch from? As I understand svntogit is just a mirror of an svn repo, and not a working one. It also contains two slightly different copies of the PKGBUILD, in two different subdirs (trunk, and repos/community-x86_64) (although I suppose
[12:32:24] <aquilairreale> they are different because asp32 patches one of them on the fly). Which one am I supposed to touch, if any?
[12:33:16] <deep42thought> you're supposed to touch none of them
[12:33:48] <deep42thought> have a look here: https://git2.archlinux32.org
[12:33:49] <phrik> Title: README.md - archlinux32/packages - Package customizations and pure-i686 packages (at git2.archlinux32.org)
[12:34:05] <deep42thought> but if that's too much for you, I'm also happy with a plain diff/patch of the PKGBUILD
[12:34:16] <deep42thought> I can translate that to our format as necessary
[12:35:39] <deep42thought> btw: the difference between trunk/ and repos/.../ is most probably already there upstream
[12:38:04] <aquilairreale> Ok, no, it's just a bit hard to come by this information but it's just what I needed
[12:38:51] <aquilairreale> E.g. first time I see git2.archlinux.org, the project's webpage only mentions git.archlinux.org
[12:39:04] <aquilairreale> Which is offline
[12:42:48] <aquilairreale> Anyway, I checked arch's svntogit web interface earlier, and it shows no difference between trunk/ and repos/, that's why I reckoned asp32 could be doing some edits to the PKGBUILD. After all, the only difference between them is the fact that the one in repos/, as pulled by asp32, has a wider array of architectures including 32 bit ones, which upstream doesn't have
[13:07:34] <buildmaster> pentium4/vulkan-tools are broken (says buildknecht2): https://archlinux32.org
[13:22:11] <buildmaster> i686/vulkan-tools are broken (says buildknecht2): https://archlinux32.org
[13:38:55] <deep42thought> aquilairreale: yes, asp32 also appends our architectures (I forgot about that)
[13:39:18] <deep42thought> git2 is not mentioned on the front page, because it is really just intended as a backup, to keep the build slaves alive in case of git being down
[13:53:44] -!- svenstaro has quit [Ping timeout: 252 seconds]
[13:54:53] <aquilairreale> Ok, it was a learning experience, hope I got it right; your system for patching upstream was all a bit new to me (with the main git being down all the time :P)
[13:55:03] <aquilairreale> Anyhow, let me know if there are any problems
[13:55:45] <deep42thought> currently, I only see one problem: I haven't gotten any email from you (yet?)
[13:55:55] <aquilairreale> Yeah, I just sent it
[13:55:57] <aquilairreale> Give it time
[13:56:00] <deep42thought> :-)
[13:56:37] <aquilairreale> No rush :) brb
[14:00:25] <deep42thought> damn, one should *never* edit routing tables from remote ...
[14:18:08] -!- waldner has joined #archlinux32
[14:19:24] <waldner> I'm trying to install wireguard-dkms and getting an error during module build: "cc1: error: incompatible gcc/plugin versions" and then "cc1: error: fail to initialize plugin ./scripts/gcc-plugins/structleak_plugin.so"
[14:19:51] -!- svenstaro has joined #archlinux32
[14:20:06] <waldner> it seems like forums and bugtracker are both down, so I'm not able to check whether it's a known thing
[14:31:47] <buildmaster> pentium4/dart is broken (says rechenknecht): https://archlinux32.org
[14:48:40] -!- abaumann has joined #archlinux32
[14:48:41] <buildmaster> Hi abaumann!
[14:48:41] <buildmaster> !rq abaumann
[14:48:43] <phrik> buildmaster: <abaumann> bugs are always smarter than you are :-)
[14:48:45] <deep42thought> Hi abaumann!
[14:48:53] <abaumann> hi deep42thought
[14:49:27] <abaumann> @archlinux32.org, ahem. guess where this mailserver is running..
[14:49:43] <deep42thought> ooooops, totally forgot *that*
[14:51:11] <aquilairreale> Oh, lol
[14:51:33] <abaumann> but we don't want to leak real email adresses here maybe.
[14:51:48] <deep42thought> np, the patch is already applied :-)
[14:52:15] <abaumann> ha. there is a secodary MX in the DNS
[14:52:25] <abaumann> pointing to a working place.
[14:52:32] <abaumann> so, it's just IMAP
[14:52:32] <deep42thought> this is my server
[14:52:40] <abaumann> being down, I meant.
[14:52:40] <deep42thought> but it simply tries to forward to tyzoids
[14:52:53] <abaumann> ah. :-)
[14:54:26] <abaumann> mmh.. my build slaves are somewhat unstable..
[14:55:09] <deep42thought> if it was around 14 o'clock (CET), then it might just have been me fucking up the routing table of git2.archlinux32.org
[14:55:27] <abaumann> ah :-)
[14:55:39] <abaumann> no, they are longer down than that..
[14:55:52] <deep42thought> I had ipv6 trouble before
[14:56:13] <abaumann> yeah mine look like crashes of some sort..
[14:56:26] <abaumann> I was thinking to make this dynamic:
[14:56:33] <abaumann> check the workload on https://buildmaster-status.archlinux32.org
[14:56:33] <phrik> Title: Buildmaster for Archlinux32 packages (i686, pentium4, any) (at buildmaster-status.archlinux32.org)
[14:56:45] <abaumann> then add slaves, if the number of tasks goes over certain levels..
[14:57:07] <abaumann> btw, I don't know it I already mentioned that: https://archlinux32.org is stale
[14:57:09] <phrik> Title: Buildmaster for Archlinux32 packages (i486, i686, pentium4, any) (at archlinux32.org)
[14:57:21] <abaumann> or totally not showing the same thing..
[14:57:37] <deep42thought> hmm, interesting
[14:58:08] <abaumann> the list of build slaves are both showing "last connections" from today.
[14:58:14] <abaumann> so, something _is_ synching
[14:58:36] <deep42thought> the webdir on buildmaster-status is still not a git, right?
[14:58:39] <abaumann> aha. the archlinux32-keyrings hit the slaves..
[14:58:44] <abaumann> no,
[14:58:52] <deep42thought> keyring should be fine by now
[14:58:53] * abaumann remebers he had a todo there..
[14:59:11] <abaumann> yeah. but when installing a new chroot, I get into key errors.
[14:59:14] <abaumann> on some slaves.
[14:59:49] <deep42thought> with archlinux32-keyring-20191103-1.0-any.pkg.tar.xz ?
[15:00:08] <abaumann> I have to debug, something is weird. I suspect a wrong stale mirror of sorts.
[15:00:17] <deep42thought> might be
[15:00:34] <deep42thought> btw: any luck with the redundand bd?
[15:02:55] <abaumann> not yet..
[15:03:03] <abaumann> I prepared two test servers to play with..
[15:06:52] <deep42thought> the difference between status-buildmaster and /buildmaster on the front page is, that the latter uses a newer version of the files which includes statistics for i486
[15:07:06] <deep42thought> so: everything is nominal
[15:10:41] <aquilairreale> deep42thought: Ok, I see the patch is online, thank you for your time
[15:10:46] <aquilairreale> Gotta go now, cya
[15:10:53] <deep42thought> cu
[15:10:55] -!- aquilairreale has parted #archlinux32
[15:13:19] <abaumann> aha. so, all ok :-)
[15:13:29] <deep42thought> I updated the files (for now)
[15:13:37] <deep42thought> but we should really get this version-controlled ;-)
[15:26:15] <deep42thought> gotta go - cu later!
[15:26:20] -!- deep42thought has quit [Quit: Leaving.]
[15:34:45] <buildmaster> pentium4/kalzium is broken (says eurobuild6-3): https://archlinux32.org
[15:41:10] -!- abaumann has quit [Quit: leaving]
[15:53:05] -!- thePiGrepper has quit [Ping timeout: 276 seconds]
[15:53:14] -!- thePiGre1per has joined #archlinux32
[16:05:43] <buildmaster> pentium4/kio-extras are broken (says eurobuild6-3): https://archlinux32.org
[16:05:51] T`aZ is now known as T\x60aZ
[16:19:31] <buildmaster> pentium4/lokalize is broken (says eurobuild1): https://archlinux32.org
[16:20:27] <buildmaster> i686/dart is broken (says eurobuild6-1): https://archlinux32.org
[16:22:46] <waldner> forgot to mention, I'm on pentium4 arch
[16:25:16] -!- MrBIOS has quit [Quit: MrBIOS]
[16:52:26] <buildmaster> i686/kalzium is broken (says eurobuild6-3): https://archlinux32.org
[17:06:42] -!- MrBIOS has joined #archlinux32
[17:18:13] -!- abaumann has joined #archlinux32
[17:18:13] <buildmaster> Hi abaumann!
[17:18:13] <buildmaster> !rq abaumann
[17:18:15] <phrik> buildmaster: <abaumann> well, we found out that those disks _ARE_ actually slow disks then :-)
[17:18:39] -!- deep42thought has joined #archlinux32
[17:18:40] <buildmaster> Hi deep42thought!
[17:18:40] <buildmaster> !rq deep42thought
[17:18:42] <phrik> buildmaster: <deep42thought> otoh I have the feeling, archlinux32 dies anyway if I get overrun by a bus
[17:18:43] <abaumann> deep42thought: found stale archlinux32-keyrings on the hosts where the build slaves run on, so the initial checkout was failing.. should all be fixed now.
[17:19:08] <buildmaster> i686/kio-extras are broken (says eurobuild6-2): https://archlinux32.org
[17:19:39] <deep42thought> abaumann: I just thought some more about the distributed block device
[17:19:52] <abaumann> ok.
[17:19:57] <deep42thought> we could even set up a separate vpn (tinc) for the replicates and do everything over it
[17:20:03] <deep42thought> this would simplify mysql, too
[17:20:14] <deep42thought> and we could add more machines (behind nat-routers)
[17:20:27] <abaumann> I'm just pretty sure most things work in LANs, but not in WANs
[17:20:36] <deep42thought> and I think, the asynchronous variant should be sufficient - a fsck in case of failure seems ok
[17:20:47] <abaumann> that's also my opinion
[17:20:51] <deep42thought> what makes it fail in wans?
[17:20:57] <abaumann> setup?
[17:20:57] <deep42thought> the latency?
[17:21:04] <abaumann> maybe it can be tunnelled..
[17:21:08] <deep42thought> tinc
[17:21:12] <deep42thought> ^ is nice :-)
[17:21:18] <abaumann> ah. yes. I remember.
[17:21:26] <abaumann> Have to have a closer look at tinc
[17:30:46] <deep42thought> it's basically a decentralized, encrypting vpn
[17:30:56] <abaumann> better than N*VPN :-)
[17:32:03] <deep42thought> and it uses udp-hole-punching to make direct connections (for data only)
[18:55:13] <deep42thought> abaumann: I would move this topic a little forward today/tomorrow/this weekend by installing tinc on the buildmaster, archlinux32.org and my server and doing the database replication over that
[18:55:34] <deep42thought> then it's also easier for you to test the block device over tinc
[18:59:20] <abaumann> yes, that sounds good.
[18:59:38] <abaumann> there is one problem: the block devices should use the raw disk partition..
[18:59:49] <deep42thought> loopdevice?
[18:59:50] <abaumann> ..which currently runs a system :-)
[19:00:00] <abaumann> yeah. maybe.
[19:00:29] <abaumann> currently I'm learning what I missed the passed N years in terms of redundancy and high availability..
[19:00:36] <deep42thought> loopdevice on luks on lvm on another nbd :-)
[19:00:57] <abaumann> I actually wonder if you can do LVM on those replicas..
[19:01:09] <deep42thought> I would think so
[19:01:21] <abaumann> would be extremly helpful
[19:01:29] * deep42thought has never done LVM
[19:02:09] <deep42thought> I'm usually doing the other extreme: using luks on raw disks and putting a fs directly inside the luks - no partition tables involved
[19:02:20] <abaumann> I'm still confused how you stack things: RAID/mdadm, then lvm, then DRDB, then ext4
[19:02:28] <abaumann> ah. OpenStack.
[19:02:32] <abaumann> hence the name :-)
[19:02:44] <deep42thought> cannot you run drdb on any device?
[19:02:52] <abaumann> on any block device, yeah.
[19:02:54] <abaumann> pretty much.
[19:03:09] <abaumann> the question is performance and how easy you can shoot in your own foot.
[19:03:12] <deep42thought> well, then why the hassle with another raid underneath? or lvm?
[19:03:32] <abaumann> so a local RAID serves for additional local hard disk redundancy
[19:03:33] <deep42thought> well, raid1 might be nice (but may be at the bottom already, anyways)
[19:03:41] <abaumann> LVM is just there to increase partitions when they are full
[19:03:53] <deep42thought> why do I need lvm for that?
[19:04:14] <abaumann> you cannot change the size of the hard disk on the fly?
[19:04:15] <deep42thought> can't I increase the raid disk size and everything scales as usual?
[19:04:22] <abaumann> ah. yeah.
[19:04:34] <deep42thought> I have to go now, sry
[19:04:36] <abaumann> I was thinking to closely. I usually have RAID 1 of two disks.
[19:04:40] <abaumann> ok. np
[19:04:41] <abaumann> cu
[19:04:47] <deep42thought> and if you use larger disks, it will scale
[19:04:56] <deep42thought> and becaus it's raid1, you can do it live
[19:05:15] <deep42thought> or do you expect to hit the maximum capacity of a single disk?
[19:05:23] <deep42thought> anyways: cu!
[19:05:26] -!- deep42thought has quit [Quit: Leaving.]
[19:09:20] * abaumann is sorry, had to fetch a boiling tea
[19:09:41] <abaumann> exactly that. hitting the physical limit of the disk.
[19:09:55] <abaumann> ok, a RAID10 should help then..
[19:10:06] -!- MrBIOS_ has joined #archlinux32
[19:10:13] <abaumann> for the buildmaster it means 'migrations'..
[19:10:18] <abaumann> ..and I don't like migrations.
[19:10:57] -!- MrBIOS has quit [Ping timeout: 240 seconds]
[19:10:57] MrBIOS_ is now known as MrBIOS
[19:33:39] -!- abaumann has quit [Quit: leaving]
[19:43:00] -!- davor has quit [Ping timeout: 265 seconds]
[19:45:14] -!- davor has joined #archlinux32
[19:59:52] -!- urraka has joined #archlinux32
[20:05:21] <urraka> hi
[20:06:32] <urraka> i'm getting some signature errors in pacman
[20:06:57] <urraka> error: libmad: signature from "Andreas Baumann (sign) <mail@andreasbaumann.cc>" is unknown trust
[20:07:02] <urraka> whats up with that
[20:19:11] -!- samantaz_ has joined #archlinux32
[20:21:54] -!- abaumann has joined #archlinux32
[20:21:54] <buildmaster> Hi abaumann!
[20:21:54] <buildmaster> !rq abaumann
[20:21:56] <phrik> buildmaster: <abaumann> don't make serious jokes ;-)
[20:22:22] <abaumann> urraka: hi, try to update only the keyring first with 'pacman -Sy archlinux32-keyring'
[20:22:37] <abaumann> then the usual 'pacman -Syyu'
[20:23:16] <urraka> abaumann: i tried pacman-key --refresh-keys and that seemed to fix it
[20:23:24] <abaumann> that's the other option, yes.
[20:23:33] <urraka> cool
[20:23:36] <abaumann> if you have access to a key-sever that is :-)
[20:24:04] <urraka> i guess i have
[20:24:08] <urraka> thanks still
[20:24:11] <abaumann> np
[20:40:35] -!- abaumann has quit [Quit: leaving]
[20:56:17] -!- isacdaavid has joined #archlinux32
[21:00:13] <buildmaster> pentium4/dmd is broken (says nlopc46): https://archlinux32.org
[21:45:01] <buildmaster> pentium4/ldc is broken (says nlopc46): https://archlinux32.org
[21:56:47] -!- isacdaavid has quit [Quit: Leaving.]
[21:56:51] <buildmaster> pentium4/mustache-d is broken (says nlopc46): https://archlinux32.org
[22:52:30] <buildmaster> any/firefox-tridactyl is broken (says nlopc46): https://archlinux32.org
[22:57:32] <buildmaster> i686/ldc is broken (says nlopc46): https://archlinux32.org
[23:19:12] <buildmaster> i686/dmd is broken (says nlopc46): https://archlinux32.org
[23:19:19] -!- MrBIOS has quit [Read error: Connection reset by peer]
[23:21:25] -!- MrBIOS has joined #archlinux32
[23:26:21] <buildmaster> i686/mustache-d is broken (says nlopc46): https://archlinux32.org
[23:58:29] -!- urraka has quit [Quit: WeeChat 2.6]
[23:59:59] -!- MrBIOS_ has joined #archlinux32