00:38:34  * runciter7joined
00:41:25  * runciter7quit (Remote host closed the connection)
01:19:19  * k1ll9joined
01:24:02  * k1ll9quit (Remote host closed the connection)
03:16:59  * node-ghjoined
03:16:59  * node-ghpart
05:10:33  * thefourtheyejoined
07:49:23  * Nirkus20joined
07:52:07  * Nirkus20quit (Remote host closed the connection)
09:09:53  * sgimenojoined
09:43:52  * doom18joined
09:46:46  * doom18quit (Remote host closed the connection)
09:59:56  * sxajoined
12:15:54  * bootkiller4joined
12:17:55  * bootkiller4quit (Killed (Sigyn (Spam is off topic on freenode.)))
12:25:15  * Kraps8joined
12:28:03  * Kraps8quit (Remote host closed the connection)
13:52:09  * node-ghjoined
13:52:09  * node-ghpart
13:52:15  * mmarchini_part
13:52:48  * mmarchinijoined
13:54:12  <mmarchini>is there any easy way to duplicate a job configuration? I want to take a look into https://github.com/nodejs/build/issues/746, but I don't want to mess with our node-test-pull-request job
13:58:30  <refack>mmarchini: https://ci.nodejs.org/job/mmarchini-node-test-pr/configure
13:59:24  <refack>and https://ci.nodejs.org/job/mmarchini-node-test-pr-lite/configure
13:59:36  <mmarchini>thanks! did you create it or is it created automagically by jenkins?
14:00:17  <refack>Jenkins allows you to create a new item by duplicating an exsisting item (it's a shallow duplicate, so the sub jobs __are__ shared)
14:01:09  <refack>I did some work WRT to pinging, using the GHPR plugin (https://wiki.jenkins.io/display/JENKINS/GitHub+pull+request+builder+plugin)
14:03:42  * pskosinski23joined
14:04:08  <refack>It's in https://ci.nodejs.org/job/refack-test-pull-request-lite-pipeline/configure that watches https://github.com/refack/node/pull/6
14:04:30  <mmarchini>Cool! I think we can start with something simpler though (just an e-mail notification field on the job + a E-mail Notification Post Build Action), and have something more sophisticated in future versions
14:05:01  <refack>We also have a mailer plugin, if it can take parameters from the job, that should be easy
14:06:14  <refack>https://ci.nodejs.org/job/node-test-commit-v8-linux/configure (at the bottom)
14:08:22  * pskosinski23quit (Remote host closed the connection)
14:09:50  <mmarchini>Not sure if this plugin is enough though, we can't configure it to send after every build (it only sends on failed builds or successful builds after failed ones)
14:11:01  <refack>It might expose a groovy api
14:11:50  <refack>IMHO adding a comment to the PR thread, would be optimal, e-mail might be noisy
14:41:00  * sadmac29joined
14:41:35  * sadmac29quit (Remote host closed the connection)
14:57:48  * node-ghjoined
14:57:48  * node-ghpart
15:04:34  * dwfreed18joined
15:07:35  * seishunjoined
15:07:50  <seishun>what's going on here? https://ci.nodejs.org/job/node-test-commit-linux/nodes=ubuntu1404-32/21601/console
15:08:02  <mmarchini>I guess the best way to achieve this is to add a checkbox (Comment on PR when build finishes) which will make github-bot will comment with CI result on the PR. This way we only need to add another parameter to `node-test-pr`, propagate this parameter to `post-build-status-update` which will forward it to the bot's API. Only code changes we need to make are on `jenkins-status.js` to comment on the PR when the build ends.
15:08:25  <mmarchini>refack: does that make sense?
15:10:26  <mmarchini>or we can use this: https://wiki.jenkins.io/display/JENKINS/Email-ext+plugin
15:11:54  * dwfreed18quit (Ping timeout: 252 seconds)
15:23:28  * fwind77joined
15:27:49  <refack>I'm a bit conflicted. I don't want to ~write~ maintain more code. I wish we could reuse existing code...
15:28:11  * fwind77quit (Remote host closed the connection)
15:28:40  <refack>But it does seem like getting the bot to comment is low hanging fruit
15:29:44  <refack>BTW I wouldn't propagate it. Just have one hooked point
15:31:57  <refack>BTW 2 we have a plugin that can run arbitrary groovy on certain events, like job completed, and it can check if it's a parent job
15:34:10  <refack>Confluence Mobile - Jenkins Wiki (https://wiki.jenkins.io/plugins/servlet/mobile?contentId=82020523#content/view/82020523)
15:36:43  <joyee>On the upside if we can put the code on github we may get more help maintaining them
15:44:57  <refack>If we could write Jenkins plugins in JS, that might be great. It has all the APIs just no easy way to script them (in this context Java is not easy)
15:46:10  <joyee>you mean remote APIs through HTTP endpoints, or like plugins installable on the workers?
15:49:07  <refack>Maybe we embed rhino. Local code that uses Jenkins APIs, for example `onJobEnd(if job.isParent {postGitHubComment(job.resault, job.url, job.failedTests)})`
15:49:45  <joyee>is rhino even a thing?
15:49:57  <joyee>(still?)
15:50:13  <refack>I've heard it had a small revival
15:51:47  <refack>https://github.com/mozilla/rhino/pulse/monthly
15:51:59  <joyee>It seems to involve more work than just talking to the bot..but then the bot is not in a great place in terms of maintenance either
15:52:18  <tniessen>ARM seems to be broken?
15:52:57  <tniessen>See e.g https://ci.nodejs.org/job/node-test-binary-arm/3400/
15:53:14  <tniessen>And this https://ci.nodejs.org/job/node-test-pull-request/17243/
15:53:43  <joyee>`failed to remove out/Release/.nfs0000000000c2d81c00000c87`
15:54:04  <joyee>I recall seeing this before...? refack
15:54:13  <refack>That means there's a zombie node
15:54:15  <joyee>the pies
15:54:45  <joyee>so we should just log in and kill them?
15:54:56  <refack>A zombie holds the binary file open, and if it's NFS you can't delete it
15:55:00  <refack>Yes
15:55:50  <refack>I'm AFK
15:55:51  <joyee>eh maybe I should pattern match and prompt fixes in ncu-ci..
15:56:36  <refack>(I need to set up the ssh config on my phone 🙄)
15:58:26  <joyee>https://www.irccloud.com/pastebin/PtZomu9j/
15:58:32  <joyee>I tried to ssh but got this..
15:58:58  <joyee>on `test-requireio_continuationlabs-debian9-armv6l_pi1p-1`
16:02:58  <refack>So could you set the node as offline?
16:03:06  <refack>In Jenkins
16:03:47  <joyee>Looks like there are many of them
16:04:10  <joyee>Not sure if I should take them all down..or just some of them?
16:06:33  <joyee>OK not all of them are nfs problems
16:06:43  <joyee>https://ci.nodejs.org/job/node-test-binary-arm/3401/RUN_SUBSET=5,label=pi3-docker/console this is a failed test
16:06:55  <joyee>Anyway I marked test-requireio_continuationlabs-debian9-armv6l_pi1p-1 offline
16:07:16  <joyee>There is this mysterious `/tmp/jenkins1055219543608517429.sh: line 41: /home/iojs/.ssh/known_hosts: No such file or directory`
16:07:20  <joyee>https://ci.nodejs.org/job/node-test-binary-arm/3400/RUN_SUBSET=3,label=pi1-docker/console
16:08:27  <refack>long long list of live node procs
16:08:46  <joyee>?
16:08:58  <refack>big trouble
16:09:06  <refack>Weird situation
16:09:26  <joyee>I think I walked the CI yesterday and have seen several of this
16:10:28  <joyee>https://www.irccloud.com/pastebin/j37lpW6K/
16:10:31  <joyee>Oh yeah..
16:10:35  <joyee>it's a long list
16:11:23  <joyee>Only failed 2 PRs though
16:12:07  <joyee>Why is the known_hosts not there though
16:12:51  <refack>Yay I got to ssh from my phone
16:13:03  <joyee>wow
16:13:08  <refack>test-requireio_continuationlabs-debian9-armv6l_pi1p-1 did not have `.ssh` directory
16:13:14  <joyee>eh?
16:14:01  <refack>it (as iojs) needs to ssh to to fetch cross-compiled binaries
16:14:16  <refack>I'll check if that solve the problems later
16:15:06  <joyee>Where does it get the ssh config? The playbook? (sorry super noob here)
16:16:46  <refack>Jenkins injects it (into an ssh-agent)
16:17:21  <refack>It's quite nice, I still havn't see it leak the key
16:18:41  <refack>seishun: the git repository got stuck in a previus operation, and left `./git/lock` behind. I deleted it.
16:18:56  <joyee>wrong channel?
16:18:58  <refack>The interesting part is why are we testing node11 on x86 linux
16:19:03  <joyee>oh wait
16:19:05  <joyee>right channel
16:19:12  <refack>yeah (back log)
16:23:53  <refack>joyee: on a different topic, do you know if I can get IR/AST from GN,
16:24:09  <joyee>Um, you mean parsing GN?
16:24:18  <refack>I want to try a GN to GYP transformer
16:24:36  <joyee>There is a gn plugin in vscode
16:24:41  <refack>Exactly
16:24:43  <joyee>Not sure what it is using
16:25:41  <refack>Hmm, probably just syntax coloring and forking the exe, buy it might hold the grammar
16:25:47  * seishunquit (Ping timeout: 240 seconds)
16:25:53  <joyee>Maybe there is something in chromium
16:26:21  <refack>I'll see if I can break the GN exe to parts
16:27:32  <refack>I'm at a point where I have the GYP AST, so I could write GYP to GN, but I don't like GN
16:27:45  <refack>I have great ideas for GYP
16:28:03  <refack>Like linker cache
16:29:02  <joyee>Still not sure what the long-term plan should be..
16:29:23  <joyee>The CMake path is as shaky as GYP for us
16:30:46  <joyee>eh looks like the parser is hand-written in C++
16:32:25  <refack>In my POV they are all shaky. But GYP I know, and I like python 🤷
16:33:14  <refack>The llvm people don't like actually building and distributing binaries
16:36:49  * seishunjoined
16:52:19  * mowntan1joined
16:56:03  <tniessen>refack: seems like CI isn't stopping once again : https://ci.nodejs.org/job/node-test-commit-linux/21602/
16:56:19  * sxaquit (Ping timeout: 244 seconds)
16:56:27  <tniessen>Do I need to restart the whole CI? ...
16:57:30  * mowntan1quit (Ping timeout: 264 seconds)
18:16:03  * toddf16joined
18:21:03  * toddf16quit (Ping timeout: 244 seconds)
20:17:52  * late-13joined
20:20:33  * late-13quit (Remote host closed the connection)
20:25:16  * seishunquit (Ping timeout: 246 seconds)
20:28:12  * Guest89797joined
20:30:37  * Guest89797quit (Remote host closed the connection)
20:38:44  * Welcomejoined
20:39:31  * Welcomequit (Remote host closed the connection)
21:01:34  * seishunjoined
21:10:04  * seishunquit (Ping timeout: 240 seconds)
21:15:24  * b225joined
21:15:48  * kenlumbo15joined
21:18:39  * kenlumbo15quit (Remote host closed the connection)
21:20:48  * b225quit (Ping timeout: 272 seconds)
21:30:12  * node-ghjoined
21:30:12  * node-ghpart
21:30:22  * node-ghjoined
21:30:22  * node-ghpart
21:31:34  * node-ghjoined
21:31:34  * node-ghpart
23:42:38  <tniessen>ubuntu1404-32 is failing with weird infra issues: *** Error in `git': double free or corruption (out): 0x099e9868 ***
23:42:49  <tniessen>see https://ci.nodejs.org/job/node-test-commit-linux/21625/nodes=ubuntu1404-32/console
23:42:53  <tniessen>and arm still seems to be down
23:43:43  <refack>I think I saw a build finishing on ARM
23:43:54  <refack>Are you trying to resume?
23:44:10  <refack>You can't resume an arm-fanned
23:44:36  <tniessen>Oh... So do I have to restart the whole CI for that?
23:45:12  <tniessen>And node-test-commit-linux seems to be very busy? I started a couple of builds two hours ago and they are still pending
23:50:36  * blu3r4d0n2joined
23:53:09  * blu3r4d0n2quit (Remote host closed the connection)
23:56:01  <refack>Two platforms are really slow ubuntu14 and 17
23:56:26  <tniessen>are we using some kind of ccache on CI across runs?
23:56:46  <tniessen>(out of curiosity, looking for ways to speed things up)
23:57:02  <tniessen>got to go to sleep now, but thanks for the info refack! :)