00:06
<
gchristensen >
samueldr: do you have a PR I can see which broken tested?
00:06
<
gchristensen >
(ie, made it disappear)
00:07
<
{^_^} >
#57557 (by matthewbauer, 5 weeks ago, merged): nixos/virtualbox: add swap file
00:07
<
samueldr >
(which is why I circled back to my PR, it hopefully would have made ofborg frown)
00:19
<
gchristensen >
I think I need some sort of /tmp cleaner
00:22
<
gchristensen >
not sure this is what I want?
00:23
<
gchristensen >
310GB used up in /tmp
00:23
<
samueldr >
Example 4. Apply clean up during boot and based on time
00:23
<
samueldr >
the based on time seems likely
00:24
<
gchristensen >
$ sudo rm -rf ./nix-build-*
00:24
<
samueldr >
though not sure you want to nuke /tmp as some software might still be using it for shared tokens during a boot, expecting it won't get nuked
00:24
<
gchristensen >
sudo: unable to execute /nix/store/68z2cvbzws1pn0z8dhgfkmws75r2z7gm-coreutils-8.29/bin/rm: Argument list too long
00:25
<
samueldr >
can nix be told to build into another directory than /tmp directly? e.g. /tmp/nix/nix-build-* so that it's easier to tell systemd to clean that directory wholesale?
00:25
<
samueldr >
though, it looks like it handles globs in its syntax
00:25
<
gchristensen >
not sure
00:29
<
gchristensen >
find . -maxdepth 1 -name 'nix-build-*' -print0 | xargs -0 -n 10 sudo rm -rf
00:29
<
gchristensen >
always thrilling
00:33
<
gchristensen >
looks like "R /tmp/nix-build-* - - - 1d -" seem right samueldr
00:33
<
samueldr >
let's hope, for many reasons, that there won't be a build longer than one day!
00:33
<
gchristensen >
hehe
00:34
<
samueldr >
the PR checks passed; but it looks like the tested one isn't there?
00:35
<
gchristensen >
I haven't deployed yet
00:35
<
gchristensen >
deploy failed because I ran out of space
00:51
<
gchristensen >
https:?/butt.holdings/
00:51
<
gchristensen >
let's see
00:56
<
gchristensen >
is the "@grahamcofborg" prefix necessary, what if we just dropped it or replaced it with /build and /eval or something
00:57
<
samueldr >
not necessary, slash commands would work fine
00:58
<
gchristensen >
hmm okay
00:58
<
samueldr >
as long as you can receive them
00:58
<
gchristensen >
b/c @grahamcofborg has been not really necessary for a long time now
00:58
<
samueldr >
though it probably should handle it for a sunsetting period...
00:59
<
gchristensen >
yeah we could do that just fine
00:59
<
{^_^} >
#59994 (by grahamc, 52 minutes ago, open): (do not merge) Revert "nixos/virtualbox: Fixes configuration to evaluate" to ensure ofborg checks this properly
00:59
<
samueldr >
though, only orthogonal, there needs to be a better kind of feedback for ofborg, that something is happening, I don't know what
01:00
<
samueldr >
hopefully that's the last PR that disappears tested
01:01
<
samueldr >
yay, that one is almost silly
01:36
<
infinisil >
Well, I finally got a logging entry of these rabbitmq failures my nixbot has been seeing
01:36
<
infinisil >
Apr 22 03:24:21 protos nixbot[3834]: AMQP channel exception: Terminated True "received fatal error: UnexpectedMessage" (Error_Protocol ("remote side fatal error",True,UnexpectedMessage))
01:37
<
gchristensen >
interesting
01:37
<
infinisil >
That's 12 minutes ago, 6 seconds after worldofpeace did the ,locate in #nixos
01:38
<
infinisil >
I wish the error wasn't this vague..
01:40
<
gchristensen >
what are you using for your client?
01:44
<
gchristensen >
interestign
01:44
<
gchristensen >
I can't find "UnexpectedMessage" "Error_Protocol" or "remote side fatal error" in that lib or rabbitmq
01:47
<
infinisil >
The "AMQP channel exception" is from me at least
01:47
<
gchristensen >
Apr 22 01:24:21 zoidberg rabbitmq-server[16478]: TLS server: In state connection at ssl_connection.erl:1020 generated SERVER ALERT: Fatal - Unexpected Message - {unexpected_msg,{alert,2,20,
01:48
<
gchristensen >
Apr 22 01:24:21 zoidberg rabbitmq-server[16478]: {"ssl_cipher.erl",304},
01:48
<
gchristensen >
Apr 22 01:24:21 zoidberg rabbitmq-server[16478]: undefined,decryption_failed}}
01:48
<
gchristensen >
"closing AMQP connection: {inet_error,{tls_alert,"unexpected message"}}"
01:48
* infinisil
digs into Haskell TLS library sources
01:49
<
infinisil >
Error_Protocol and UnexpectedMessage sure is from that one
01:49
<
infinisil >
remote side fatal error too
01:52
<
infinisil >
Really weird
03:16
orivej has quit [Ping timeout: 268 seconds]
07:30
orivej has joined #nixos-borg
07:45
orivej has quit [Ping timeout: 246 seconds]
07:47
orivej has joined #nixos-borg
07:52
orivej has quit [Ping timeout: 250 seconds]
08:52
orivej has joined #nixos-borg
10:44
orivej has quit [Ping timeout: 246 seconds]
10:46
orivej has joined #nixos-borg
11:02
orivej has quit [Ping timeout: 255 seconds]
11:32
orivej has joined #nixos-borg
12:05
orivej has quit [Ping timeout: 245 seconds]
12:22
orivej has joined #nixos-borg
12:57
orivej has quit [Remote host closed the connection]
12:58
orivej has joined #nixos-borg
13:53
<
{^_^} >
[ofborg] @grahamc pushed to revert-353-revert-351-stats « nixenv: Return a different error if stats parse fails »:
https://git.io/fj3sd
14:59
<
gchristensen >
samueldr: with the added nixos check, might be stretching eval time a bit
14:59
<
gchristensen >
might be time to upgrade to bigger evaluators
15:10
jtojnar has joined #nixos-borg
15:14
jtojnar has quit [Client Quit]
15:59
orivej has quit [Ping timeout: 250 seconds]
17:29
orivej has joined #nixos-borg
18:27
orivej has quit [Ping timeout: 250 seconds]
20:33
orivej has joined #nixos-borg
20:50
<
{^_^} >
#59820 (by vbgl, 4 days ago, open): coqPackages.tlc: 20180316 -> 20181116
20:50
<
infinisil >
It's still pending, but nobody told it to eval since 4 days at least
20:51
<
gchristensen >
nice
20:51
<
gchristensen >
oops
20:51
<
gchristensen >
was scrolled up
20:52
<
gchristensen >
hmm I'm not sure
20:56
<
gchristensen >
fwiw though the queue is significantly backed up right now
20:56
<
gchristensen >
bringing up additional capacity
21:01
<
gchristensen >
these servers are so much faster than the hetzner cloud machines
21:03
kalbasit has quit [Read error: Connection reset by peer]
21:06
<
gchristensen >
Apr 22 21:04:11 f228c1ee.packethost.net d5a15bwbydg2w1r1rhdy8vlshp631q2m-unit-script-ofborg-evaluator-start[850]: INFO:ofborg::tasks::evaluate: Working on 60051
21:08
kalbasit has joined #nixos-borg
21:11
<
gchristensen >
Apr 22 21:10:56 f228c1ee.packethost.net d5a15bwbydg2w1r1rhdy8vlshp631q2m-unit-script-ofborg-evaluator-start[850]: INFO:ofborg::tasks::evaluate: done!
21:11
<
gchristensen >
7min, wow
21:11
<
gchristensen >
it is a dream
21:13
<
infinisil >
How fast were the hetzner machines usually/
21:13
<
gchristensen >
they seem like they take 20min now with the various evaluations
21:16
<
infinisil >
Oh, then 7min is indeed very impressive
21:18
<
gchristensen >
the steep drop-off when I added more evaluators ain't no coincidence :P
21:59
samueldr has quit [Ping timeout: 246 seconds]
22:02
samueldr has joined #nixos-borg
22:07
kalbasit has quit [Ping timeout: 276 seconds]
22:08
kalbasit has joined #nixos-borg
22:16
andi- has quit [Ping timeout: 250 seconds]
22:16
cransom has quit [Ping timeout: 250 seconds]
22:17
cransom has joined #nixos-borg
22:18
andi- has joined #nixos-borg
22:31
orivej_ has joined #nixos-borg
22:35
orivej has quit [Ping timeout: 245 seconds]