#archlinux32 | Logs for 2020-09-07
Back
[00:49:30] <buildmaster> i486/croc is broken (says nlopc46-i486bs0): https://archlinux32.org
[02:47:51] <buildmaster> i486/broadcom-wl is broken (says nlopc46-i486bs1): https://archlinux32.org
[05:57:11] <trotz> 2020/09/07 05:56 OK buildmaster OS updates 0 updates, 3 ignored
[08:16:13] <girls> ahuillet: I also see this on archlinuxarm, and I don't have any clues, how to solve this
[08:16:39] <girls> abaumann suggested downgrading pam - but I'm not sure, whether this solves the sshd issue, too, or if it was for something different ...
[08:50:48] <bill-auger> FWIW i just found another PAM related bug
[08:52:53] <bill-auger> on the i686 LiveISO i have made since yesterday (ive made several now, trying to solve this) there is an intermittent bug where the login takes about 30 seconds
[08:55:02] <bill-auger> it is apparently some race condition - on about 1 of each 5 boots, it logs in immediately as expected
[08:56:35] <bill-auger> i can give you guys the ISO, if that may help find the problem - i do not see high CPU; but all of the latest PAM problems are like related
[08:57:39] <bill-auger> i at least have a test case that is reproducibly bad 4 in 5 times, and good 1 in 5 times
[08:58:30] <ahuillet> bill-auger : that's the same problem that I'm describing I think, or very similar anyway.
[08:59:27] <bill-auger> i took the logs for comparrison - the most obvious difference is this error:
[08:59:29] <bill-auger> login[377]: pam_systemd(login:session): Failed to create session: Input/output error
[08:59:58] <bill-auger> thats when it takes 30 extras seconds to login - when it logs in smoothly, that error is not present
[09:00:36] <ahuillet> yes, I observed that too.
[09:00:55] <ahuillet> but then once it logs in I think I still have sshd @100% CPU or something. so seems like it might have hte same root cause
[09:00:56] <bill-auger> ahuillet: try it 10 times - you may see that it works correctly once of twice
[09:03:15] <ahuillet> yes it does
[09:05:18] <sunshavi> I have an up to date linuxarm without that issue
[09:05:38] <sunshavi> s/linuxarm/archlinuxarm/
[09:12:37] <ahuillet> yes, but that doesn't prove that there's isn't a bug
[09:35:30] <girls> sunshavi: I have those, too
[09:35:39] <girls> but I also have archlinuxarm which expose this bug
[09:45:36] <sunshavi> girls: that's weird.
[09:46:51] <sunshavi> local/pam 1.4.0-3; local/pambase 20200721.1-2
[10:02:46] <bill-auger> lemma guess? 32 bit has the bug 64 bit doesnt
[10:07:07] -!- sunshavi has quit [Remote host closed the connection]
[10:11:35] -!- sunshavi has joined #archlinux32
[10:30:28] <ahuillet> if no one else beats me to it, I expect to have time to spend on this Wednesday
[10:36:51] <ahuillet> perf says sshd spends its time in strcmp
[10:37:07] <ahuillet> and lots of hits in pam_systemd_home.so
[10:37:48] <ahuillet> ....which looks like a piece of shit gadget no one needs but that gets forced onto our systems?
[10:45:38] <ahuillet> I'll try to comment it out and reboot my server. what could possibly go wrong.
[10:51:39] <ahuillet> btw girls some perl packages seem to be outdated (perl-authen-sasl)
[10:52:39] -!- ahuillet has quit [Quit: ZNC - 1.6.0 - http://znc.in]
[10:55:40] -!- ahuillet has joined #archlinux32
[10:55:44] <ahuillet> OK, that seems to have solved the problem
[10:56:03] <ahuillet> /etc/pam.d has three lines referencing pam_systemd_home.so
[10:56:05] <ahuillet> comment them out
[10:58:03] <ahuillet> and let me know if it helps you
[11:02:53] <bill-auger> at one point, i created an unproviliged user without a home dir, and i noticed errors like "could not find dbus.home.service"
[11:06:45] <bill-auger> ahuillet: that tricked worked for me, for consecutive 5 logins, then one the 6th try, the screen went black, with only a blinking cursor
[11:08:34] <bill-auger> my x86_64 system does not have those home.so lines
[11:10:02] <bill-auger> that system is toast though i had to kill it
[11:12:18] -!- KeiraT has quit [Remote host closed the connection]
[11:21:52] -!- KeiraT has joined #archlinux32
[11:23:00] <bill-auger> no i was wrong about that - that was my openrc system - my arch x86_64 system has those entires
[11:23:31] <bill-auger> ahuillet: do you know why those entries begin with a '-' hyphen ?
[11:23:47] <bill-auger> only those do - maybe they should not
[11:26:20] <ahuillet> I don't.
[11:26:40] <bill-auger> i think its a good fix for now
[11:27:00] <bill-auger> i can make it crash only if i exit witing 1/2 second after login
[11:27:42] <bill-auger> o/c no one is going to do that - ima try rebuilding the ISO with that change
[11:28:25] <ahuillet> I wouldn't call it a fix, but it does sound like an acceptable workaround :)
[11:28:35] <ahuillet> I guess the correct solution is to enable the systemd-homed service, but I don't want that.
[11:29:37] <bill-auger> IIRC i tried - i dont have that service
[11:32:25] <bill-auger> it is homed.service
[11:33:12] <bill-auger> systemd-homed.service
[11:34:24] <bill-auger> it did not help - the login still hangs
[13:04:41] <buildmaster> i686/cargo-tarpaulin is broken (says eurobuild6-2): https://archlinux32.org
[13:05:23] <buildmaster> pentium4/cargo-tarpaulin is broken (says eurobuild6-1): https://archlinux32.org
[13:06:50] <girls> the "-" means, it's a soft-fail, e.g. failing module does not make the login (or whatever is tried) fail
[13:07:28] <girls> the error in the log should be non-fatal
[13:07:44] <girls> I removed the plugin from /etc/pam.d/ still, because the error message was filling my logs
[13:08:15] <girls> (I have some gitolite connections every few seconds, and my server truncated log 2 days ago, because it exceeded its limit)
[13:16:53] <ahuillet> I think archlinux 64 doesn't have this pam module
[13:17:06] <ahuillet> might there be a bug in archlinux32's PAM config files that enables this when it shouldn't?
[13:17:25] <girls> no, archx86_64 has this, too
[13:18:00] <girls> check /etc/pam.d/system-auth
[13:18:06] <girls> "-auth [success=1 default=ignore] pam_systemd_home.so"
[13:18:09] <girls> and co
[13:18:34] <girls> it really should be non-fatal if systemd-homed (or whatever its name was) is not running
[13:19:01] <girls> there's even a thread in the forums about it
[13:19:05] <girls> let me search it ...
[13:20:06] <girls> https://bbs.archlinux.org
[13:20:08] <phrik> Title: [solved] pam fails to find unit dbus-org.freedesktop.home1.service / Newbie Corner / Arch Linux Forums (at bbs.archlinux.org)
[13:29:24] <ahuillet> https://github.com
[13:29:26] <phrik> Title: Infinite loop in pam_systemd_home.so · Issue #16813 · systemd/systemd · GitHub (at github.com)
[13:29:30] <ahuillet> has to be our bug?
[13:30:33] <ahuillet> does archlinux32's systemd have that fix?
[13:31:40] <girls> dunno
[13:31:49] <girls> in which version is it?
[13:32:12] <ahuillet> haha I hoped you'd know how to find out :)
[13:33:33] <ahuillet> that seems to be the commit, let's see if there's an easy way to know which tag has it https://github.com
[13:33:36] <phrik> Title: Merge pull request #16824 from keszybz/no-such-unit-error · systemd/systemd@d521e69 · GitHub (at github.com)
[13:34:12] <ahuillet> 13 days ago and the last tag was v246 on July 30th: looks like the fix isn't an any official systemd release
[13:51:54] <buildmaster> pentium4/zeromq is broken (says eurobuild6-1) - already flagged out-of-date upstream on 2020-03-18: https://archlinux32.org
[13:55:20] <buildmaster> i486/zeromq is broken (says nlopc46-i486bs0) - already flagged out-of-date upstream on 2020-03-18: https://archlinux32.org
[14:49:24] -!- KeiraT has quit [Write error: Connection reset by peer]
[14:50:00] -!- KeiraT has joined #archlinux32
[15:09:07] -!- KeiraT has quit [Remote host closed the connection]
[15:09:42] -!- KeiraT has joined #archlinux32
[16:46:27] <trotz> 2020/09/07 16:45 CRIT jeti100 Archive HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[16:46:27] <trotz> 2020/09/07 16:45 CRIT jeti100 Master Mirror HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[16:52:27] <trotz> 2020/09/07 16:51 CRIT jeti100 Archive Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[16:55:27] <trotz> 2020/09/07 16:54 CRIT jeti100 Archive Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[16:55:28] <trotz> 2020/09/07 16:54 CRIT jeti100 Master Mirror Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[16:55:28] <trotz> 2020/09/07 16:54 CRIT jeti100 Master Mirror Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[16:56:28] <trotz> 2020/09/07 16:55 OK jeti100 Archive HTTP OK: HTTP/1.1 200 OK - 821 bytes in 0.369 second response time
[16:56:28] <trotz> 2020/09/07 16:55 OK jeti100 Master Mirror HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 1.245 second response time
[17:02:28] <trotz> 2020/09/07 17:02 OK jeti100 Archive Secure HTTP OK: HTTP/1.1 200 OK - 619 bytes in 0.056 second response time
[17:05:29] <trotz> 2020/09/07 17:04 OK jeti100 Archive Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[17:05:29] <trotz> 2020/09/07 17:04 OK jeti100 Master Mirror Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[17:05:29] <trotz> 2020/09/07 17:04 OK jeti100 Master Mirror Secure HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 0.061 second response time
[17:15:30] <trotz> 2020/09/07 17:14 CRIT jeti100 Master Mirror HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[17:17:29] <trotz> 2020/09/07 17:16 CRIT jeti100 Archive Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[17:27:31] <trotz> 2020/09/07 17:26 OK jeti100 Archive Secure HTTP OK: HTTP/1.1 200 OK - 619 bytes in 0.056 second response time
[17:35:31] <trotz> 2020/09/07 17:34 OK jeti100 Master Mirror HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 0.024 second response time
[17:55:30] <trotz> 2020/09/07 17:55 CRIT jeti100 Archive Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[18:03:30] <trotz> 2020/09/07 18:03 CRIT jeti100 Master Mirror HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[18:13:32] <trotz> 2020/09/07 18:13 CRIT jeti100 Archive HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[18:15:31] <trotz> 2020/09/07 18:15 OK jeti100 Archive Secure HTTP OK: HTTP/1.1 200 OK - 619 bytes in 0.417 second response time
[18:30:32] <trotz> 2020/09/07 18:29 CRIT jeti100 Archive Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[18:37:33] <trotz> 2020/09/07 18:36 CRIT jeti100 Archive Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[18:37:34] <trotz> 2020/09/07 18:36 CRIT jeti100 Master Mirror Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[18:37:34] <trotz> 2020/09/07 18:36 CRIT jeti100 Master Mirror Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[18:57:35] <trotz> 2020/09/07 18:56 OK jeti100 Archive Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[18:57:35] <trotz> 2020/09/07 18:56 OK jeti100 Master Mirror Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[18:57:35] <trotz> 2020/09/07 18:56 OK jeti100 Master Mirror Secure HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 0.060 second response time
[19:03:33] <trotz> 2020/09/07 19:03 OK jeti100 Master Mirror HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 0.925 second response time
[19:04:34] <trotz> 2020/09/07 19:03 OK jeti100 Archive HTTP OK: HTTP/1.1 200 OK - 821 bytes in 1.562 second response time
[19:11:35] <trotz> 2020/09/07 19:10 CRIT jeti100 Archive Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[19:11:35] <trotz> 2020/09/07 19:10 CRIT jeti100 Master Mirror Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[19:11:35] <trotz> 2020/09/07 19:10 CRIT jeti100 Master Mirror Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[19:18:35] <trotz> 2020/09/07 19:17 CRIT jeti100 Archive HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[19:18:35] <trotz> 2020/09/07 19:17 CRIT jeti100 Master Mirror HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[19:21:36] <trotz> 2020/09/07 19:21 OK jeti100 Archive Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[19:21:36] <trotz> 2020/09/07 19:21 OK jeti100 Master Mirror Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[19:21:36] <trotz> 2020/09/07 19:21 OK jeti100 Master Mirror Secure HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 5.263 second response time
[19:24:10] -!- abaumann has joined #archlinux32
[19:24:10] <buildmaster> Hi abaumann!
[19:24:10] <buildmaster> !rq abaumann
[19:24:11] <phrik> buildmaster: * abaumann wonders why his silly ideas are always being implemented :-)
[19:24:14] -!- deep42thought has joined #archlinux32
[19:24:14] <buildmaster> Hi deep42thought!
[19:24:14] <buildmaster> !rq deep42thought
[19:24:15] <phrik> buildmaster: <deep42thought> I like to move it, move it; I like to move it, move it
[19:24:18] <abaumann> hi deep42thought
[19:24:27] <deep42thought> hi abaumann!
[19:24:53] <deep42thought> I'm trying to set up my update machine inside an arch vm
[19:25:07] <abaumann> "PAM starts to hit the fan." :-)
[19:25:08] <deep42thought> looks wonky on my thin downstream @home :-/
[19:25:14] <deep42thought> !grab abaumann
[19:25:15] <phrik> deep42thought: Bingpot!
[19:25:23] <deep42thought> yeah, that, too
[19:25:48] <abaumann> I had quite some trouble building ISOs, currently I don't have a running set of VMs.
[19:25:59] <abaumann> especially no i486 test vm
[19:28:29] * deep42thought checks his vm
[19:29:05] -!- isacdaavid has joined #archlinux32
[19:29:07] <deep42thought> mine is still up and running (with systemd-homed)
[19:29:25] <deep42thought> well, not "with systemd-homed", but with "unaltered /etc/pam.d/system-auth"
[19:30:21] <deep42thought> update didn't break it
[19:30:26] <deep42thought> (not [testing])
[19:31:03] <abaumann> I think I'll update one by one and make sure pam and pambase remain old
[19:31:30] <deep42thought> ah, I'm talking about ssh logins, again
[19:31:37] <abaumann> ah.
[19:31:57] <abaumann> there is also a broken Alix on the console now. pam_tally2.so insists to be able to write to some directory.
[19:48:36] <trotz> 2020/09/07 19:48 OK jeti100 Archive HTTP OK: HTTP/1.1 200 OK - 821 bytes in 0.083 second response time
[19:48:36] <trotz> 2020/09/07 19:47 OK jeti100 Master Mirror HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 9.147 second response time
[19:52:27] <deep42thought> probably, I need my own archive-server at home
[19:53:54] <abaumann> it's mainly disk?
[19:53:59] <deep42thought> yeah
[19:54:26] <deep42thought> the problem is the downlink-bottleneck in combination with seemingly too-agressive reverse-proxy timeouts
[19:54:37] <trotz> 2020/09/07 19:53 CRIT jeti100 Archive Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[19:54:37] <trotz> 2020/09/07 19:53 CRIT jeti100 Master Mirror Certificate HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[19:54:37] <trotz> 2020/09/07 19:53 CRIT jeti100 Master Mirror Secure HTTP CRITICAL - Invalid HTTP response received from host on port 443: cURL returned 28 - Timeout was reached
[19:55:13] <deep42thought> ah, btw: there might be some power outage tomorrow
[19:55:21] <deep42thought> at the master mirror
[19:55:35] <abaumann> ah. power outage. hehe/.
[19:55:41] <deep42thought> not sure, whether I recall the house numbers correctly or if it was the neighbour building
[20:00:37] <trotz> 2020/09/07 19:59 OK jeti100 Archive Secure HTTP OK: HTTP/1.1 200 OK - 619 bytes in 7.722 second response time
[20:02:37] <trotz> 2020/09/07 20:02 CRIT jeti100 Master Mirror HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[20:03:37] <trotz> 2020/09/07 20:02 CRIT jeti100 Archive HTTP CRITICAL - Invalid HTTP response received from host on port 80: cURL returned 28 - Timeout was reached
[20:04:38] <deep42thought> I don't know, why trotz complains so much - the master mirror is up and happily running ...
[20:14:39] <trotz> 2020/09/07 20:14 OK jeti100 Archive Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[20:14:39] <trotz> 2020/09/07 20:14 OK jeti100 Master Mirror Certificate OK - Certificate 'arch.eckner.net' will expire on Sat 10 Oct 2020 11:03:39 PM GMT +0000.
[20:14:39] <trotz> 2020/09/07 20:14 OK jeti100 Master Mirror Secure HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 8.422 second response time
[20:16:33] <abaumann> mmh. maybe nagios on the buildmaster has some network hickups..
[20:17:09] <abaumann> -28 curl DNS timeouts
[20:17:13] <abaumann> mmh. sounds familiar.
[20:18:39] <trotz> 2020/09/07 20:17 OK jeti100 Archive HTTP OK: HTTP/1.1 200 OK - 821 bytes in 0.084 second response time
[20:18:39] <trotz> 2020/09/07 20:18 OK jeti100 Master Mirror HTTP OK: HTTP/1.1 200 OK - 1408 bytes in 0.024 second response time
[20:19:51] <abaumann> I have the Hetzner DNS in /etc/resolv.conf. Maye they are a little bit flacky?
[20:20:29] <deep42thought> didn't we have trouble with them in the past?
[20:20:32] <deep42thought> why not use 127.0.0.1?
[20:20:43] <abaumann> is it a caching DNS? ours?
[20:20:50] <deep42thought> yeah, sure
[20:21:23] <deep42thought> but it's not a stub-resolver
[20:21:24] <abaumann> especially as the forwarders are the same :-)
[20:21:35] <deep42thought> oh, it is?
[20:21:40] <abaumann> yep.
[20:21:56] <abaumann> and I hope you are aware that this is basically the solution systemd took :-)
[20:22:07] <deep42thought> sure, I am
[20:22:15] <deep42thought> but I do not run more than 2 of those per network
[20:22:17] <abaumann> a proper bind might be more stable.
[20:22:25] <deep42thought> it's a proper bind
[20:22:31] <abaumann> I know :-0
[20:22:34] <abaumann> :-)
[20:22:37] <deep42thought> ?
[20:22:39] <deep42thought> you confuse me
[20:22:47] <abaumann> We set it up together. :-)
[20:22:55] <abaumann> on the buildmaster, I mean.
[20:23:01] <deep42thought> :-D
[20:23:03] <deep42thought> yeah
[20:23:09] <abaumann> the other ones would be spooky.
[20:23:12] <deep42thought> btw: it's *no* stub resolver
[20:23:20] <deep42thought> it's a fully-featured default-bind resolver
[20:23:35] <abaumann> seems to work fine with 127.0.0.1 in /etc/resolv.conf
[20:23:35] <deep42thought> plus authoritative for archlinux32.org
[20:23:39] <abaumann> let's keep it that way
[20:23:42] <deep42thought> :-)
[20:24:04] <abaumann> There might be startup hickups when booting and the named is not up already..
[20:24:18] <deep42thought> especially for iptables
[20:24:20] <deep42thought> :-/
[20:24:38] <abaumann> mmh. indeed.
[20:24:45] <abaumann> we can put them into /etc/hosts
[20:24:58] <abaumann> fenster.home.eckner.net
[20:24:59] <abaumann> hehe
[20:25:21] <deep42thought> the ip is somewhat dynamic, though :-/
[20:25:25] <abaumann> mmh.
[20:25:27] <abaumann> then no.
[20:25:47] <abaumann> or.
[20:25:56] <abaumann> mmh. no
[20:26:00] <abaumann> no round-robin.
[20:26:12] <deep42thought> this would only thin-out the errors :-D
[20:27:56] -!- Laibsch has joined #archlinux32
[20:31:24] <abaumann> the BFD:len error is some error message croaked onto the console.
[20:31:43] <abaumann> Berkley filters IIRC
[20:32:21] <abaumann> mmh. only on 32-bit?
[20:36:57] <Laibsch> I would like to install archlinux32 on a very old computer of mine. Via C3 CPU i586. It currently has Debian on it. I downloaded the arch32 i686 ISO und unsquashed the airootfs.sfs file. "arch-chroot squashfs-root/" then gives me an "illegal instruction" error which I assume is because the C3 is not i686 class.
[20:37:04] <Laibsch> How to proceed in this case?
[20:37:24] <deep42thought> can you post your /proc/cpuinfo?
[20:37:28] <deep42thought> flags I meant
[20:37:50] <abaumann> no cmov, I reckon
[20:37:58] <abaumann> So you are in for an i486 adventure :-)
[20:37:59] <deep42thought> ah, then you need the i486 iso
[20:38:13] <abaumann> the ISO not supporting zstd.
[20:38:15] <abaumann> ahem :-)
[20:38:37] <Laibsch> deep42thought: Sure. http://paste.debian.net
[20:38:38] <phrik> Title: debian Pastezone (at paste.debian.net)
[20:38:55] <Laibsch> yes, no cmov
[20:39:04] <Laibsch> I did not see an i486 ISO
[20:39:04] <deep42thought> use https://archive.archlinux32.org as temporary mirror
[20:39:14] <deep42thought> then you can update the iso to one which understands zst
[20:39:17] <abaumann> the iso is only on my mirror.
[20:39:18] <deep42thought> and then install normally
[20:39:29] <abaumann> https://archlinux32.andreasbaumann.cc
[20:39:30] <phrik> Title: Index of /arch486isos (at archlinux32.andreasbaumann.cc)
[20:39:43] <Laibsch> I see. Thanks, Andreas. Your work is much appreciated.
[20:39:45] <abaumann> but it has no support for zstd compressed packages.
[20:42:32] <abaumann> tons of "from arch32.mirrors.simplysam.us : Could not resolve host: arch32.mirrors.simplysam.us
[20:42:36] <abaumann> but that's not a problem
[20:42:43] <deep42thought> remove the mirror first :-D
[20:42:59] <abaumann> that's an option.. but then I'm loosing all this nice output. ;-)
[20:43:21] <abaumann> also, depending on the amount of memory you have, make a swap first
[20:43:29] <abaumann> pacman got a little bit more memory hungry lately. :-)
[20:43:43] <abaumann> :: File /mnt/var/cache/pacman/pkg/iputils-20190709-3.0-i486.pkg.tar.zst is corrupted (invalid or corrupted package (PGP signature)).
[20:43:47] <abaumann> yeah, updating the keys.
[20:45:53] <abaumann> hey, that ISO is old. It even lacks my own mirror. :-)
[20:46:03] <deep42thought> lol
[20:47:59] <abaumann> error: could not open file /mnt/var/cache/pacman/pkg/systemd-sysvcompat-245.6-8.0-i486.pkg.tar.zst: Unrecognized archive format
[20:48:03] <abaumann> yeah. that's a dead end.
[20:48:23] <deep42thought> use the archive for an intermediate update
[20:48:27] <deep42thought> or even install from that version
[20:48:30] <abaumann> archive mirror then as deep42thought said, then after finishing the install we can upgrade pacman, then the rest
[20:48:33] <deep42thought> and then do regular pacman -Syu
[20:48:55] <deep42thought> ah, sry, I thought, you had forgotten / not read
[20:49:18] <abaumann> I scrolled passed my terminal :-)
[20:49:23] <abaumann> sorry
[20:51:28] <abaumann> error: failed retrieving file 'openssl-1.1.1.g-2.1-i486.pkg.tar.zst' from archive.archlinux32.org : error setting certificate verify locations: CAfile: /etc/ssl/certs/ca-certificates.crt CApath: none
[20:51:36] <abaumann> yeah. http maybe.
[20:51:43] <abaumann> the ca-certs look way old on that ISO :-)
[20:52:01] <deep42thought> dang, shot my foot ... again: acme2certifier's whitelist *really* works - I tried and tried and tried to get a certificate for archive-server.archlinux32.org from my local acme server, but it only hands out certificates for "eckner.net" - just as configured
[20:54:47] <abaumann> that's why I hate auto-redirecting http to https :->
[20:54:49] <abaumann> Server = http://archive.archlinux32.org
[20:55:04] <deep42thought> err ,gimme a sec
[20:55:04] <abaumann> So you have to manually ca-certs on the ISO
[20:55:17] <abaumann> ah. that would be great.
[20:55:50] <abaumann> one day let's encrypt will encrypt 80% of the web.. and then go out of business.. so the internet is down. :-)
[20:56:08] <abaumann> why do people think adding single point of failures is a smart idea.
[20:56:21] <deep42thought> I do run my own acme server :-)
[20:56:28] <abaumann> smart :-)
[20:56:53] <deep42thought> try again
[20:57:16] <abaumann> ok.
[20:58:19] <deep42thought> archive-server, archive, mirror and sources should now be available on http, too
[20:58:25] <abaumann> thanks. :-)
[20:59:47] <deep42thought> np
[21:00:05] <deep42thought> it should have been this way from the beginning
[21:00:10] <deep42thought> (the mirror was, the rest not)
[21:01:16] <abaumann> mmh. I have to start over. My package database and my pgp keys are now completely confused on the RAMdisk :-)
[21:02:47] <deep42thought> my problem was, that there was not enough space on the ramdisk to update that first
[21:02:54] <abaumann> yep.
[21:03:23] <deep42thought> so I added another 20GB disk image to the vm, copied the iso root to that, chrooted, updated and installed from that
[21:04:57] <abaumann> yeah, and I updated the keys on the iso only not to be able to read the old databases from the archive :-)
[21:04:58] <deep42thought> but I think, it would work to install a 2019-12-07 version from the archive and then update the mirror list (in the chroot) and do `pacman -Syu`
[21:05:13] <abaumann> yes.
[21:05:15] <deep42thought> lol
[21:05:48] <abaumann> there is a reason why I do things without reading any documentation: you find out many more mistakes people can make.
[21:06:02] <deep42thought> !grab abaumann
[21:06:03] <phrik> deep42thought: Bazinga!
[21:06:49] <deep42thought> that's similar to having small children: I would have never figured out how to make the keyboard play music, but my kids all figured this out quite fast (without reading the guide, obviously)
[21:07:00] <abaumann> lol
[21:08:09] <deep42thought> (the last proband solved it with ~ 8 months)
[21:09:32] * abaumann wonders how many glasses of applesauce result from 30 apples..
[21:09:40] <abaumann> ..the answer is 6 :-)
[21:18:43] <abaumann> late.. /me yawns
[21:18:58] <abaumann> will continue tomorrow.
[21:19:05] <deep42thought> gnight!
[21:19:12] <abaumann> 5/6 test machines and 3/7 build slaves are back.
[21:19:13] <abaumann> cu.
[21:19:19] <abaumann> good night. :-)
[21:19:20] -!- abaumann has quit [Quit: leaving]
[21:31:38] <buildmaster> i486/picom is broken (says nlopc46-i486bs0): https://archlinux32.org
[21:32:41] -!- deep42thought has parted #archlinux32
[21:49:27] eschwartz is now known as oschwartz
[21:50:52] oschwartz is now known as eschwartz
[22:37:41] <buildmaster> i686/picom is broken (says eurobuild6-1): https://archlinux32.org
[22:38:51] <buildmaster> pentium4/picom is broken (says eurobuild6-2): https://archlinux32.org
[23:02:56] -!- samantaz_ has quit [Remote host closed the connection]
[23:03:24] -!- samantaz_ has joined #archlinux32
[23:45:43] -!- KeiraT has quit [Ping timeout: 240 seconds]
[23:48:43] -!- Laibsch has quit [Quit: Leaving.]
[23:50:34] -!- KeiraT has joined #archlinux32