00:57:23  <jbergstroem>italoacasas: pong
00:59:17  <italoacasas>jbergstroem hey, mhdawson already responded, was something related with linuxone
00:59:25  <italoacasas>thanks anyways
01:14:39  <jbergstroem>indutny: ok thanks
01:15:14  <indutny>np
01:15:18  <indutny>you're always welcome
01:54:34  <indutny>:D
02:15:14  <jbergstroem>indutny: pfff :-D
02:33:21  * not-an-aardvarkjoined
02:40:31  * node-ghjoined
02:40:31  * node-ghpart
05:57:24  * node-ghjoined
05:57:24  * node-ghpart
08:07:45  * rmgquit (Read error: Connection reset by peer)
08:08:11  * rmgjoined
08:25:17  * rmg_joined
08:26:54  * rmgquit (Read error: Connection reset by peer)
11:08:42  * not-an-aardvarkquit (Quit: Connection closed for inactivity)
11:25:11  * mylesborinsquit (Quit: farewell for now)
11:25:42  * mylesborinsjoined
11:53:28  * targosjoined
11:53:40  <targos>hi
11:54:06  <targos>both ubuntu1610-x64 nodes seem to be down
11:54:29  <targos>https://ci.nodejs.org/label/ubuntu1610-x64/
11:54:53  <targos>is it expected?
12:37:38  * orangemocha__quit (Read error: Connection reset by peer)
12:38:05  * orangemocha__joined
12:42:25  * node-ghjoined
12:42:25  * node-ghpart
12:52:56  * node-ghjoined
12:52:56  * node-ghpart
13:06:34  <jbergstroem>targos: checking
13:06:58  <jbergstroem>both java processes were dead
13:07:28  <jbergstroem>i'm starting to lean towards adding a prescritp to update the client jar; sohuld probably just wrap it in some timeout that doesn't bail
13:07:42  <jbergstroem>targos: (online again)
13:11:28  * chorrelljoined
13:16:20  * bnoordhuisjoined
13:17:14  <bnoordhuis>afternoon everyone. seems the shasums don't match the downloads anymore, https://github.com/nodejs/node/issues/11499
13:17:44  <bnoordhuis>i'm able to reproduce fwiw
13:21:47  <bnoordhuis>jbergstroem: you wouldn't happen to be around?
13:22:20  <jbergstroem>am!
13:22:21  <jbergstroem>ouch
13:22:46  <jbergstroem>so, is this similar to last time i wonder
13:24:02  <jbergstroem>(ref: https://github.com/nodejs/build/issues/586)
13:24:32  <bnoordhuis>ah, i remember that
13:24:48  <jbergstroem>mylesborins: you around?
13:24:59  <bnoordhuis>i checked some older releases at random and they still check out. so yes, plausible
13:26:17  <jbergstroem>urgh :( so we re-bake those releases and overwrite shasums, then update blog post?
13:26:23  <jbergstroem>not very nice to mirrors
13:26:32  <jbergstroem>well, real mirrors would probs work but yeah
13:27:39  <jbergstroem>or we make 6.9.6?
13:28:02  <jbergstroem>lets hope myles doesn't sleep in and irc as much as we do :| would appreciate his input
13:28:27  <bnoordhuis>v6.9.6 wouldn't really solve it, then we'd still have mismatching binaries
13:28:44  <bnoordhuis>it's affecting e.g. the latest v0.12 release too
13:30:15  <jbergstroem>sure, but just switching wouldn't really be the right thing to do either. just reading through the previous teps
13:30:17  <jbergstroem>*steps
13:30:47  * chorrellquit (Quit: My MacBook has gone to sleep. ZZZzzz…)
13:39:20  <italoacasas>This is just for v6?, v4 and v7 are clean?
13:44:35  <jbergstroem>we should probably test
13:44:58  <jbergstroem>and issue seems to be v6 and v0.12
13:45:32  <bnoordhuis>it's weird. i can reproduce with latest v0.12 but all v4, v6 and v7 releases check out so far
13:45:55  <bnoordhuis>but e.g. v0.12.17 is okay
13:46:15  <targos>jbergstroem: thank you
13:47:09  <jbergstroem>the shasum's seems to be touched 1 hour after last upload
13:47:14  <jbergstroem>for v6.9.5
13:47:21  * chorrelljoined
13:47:28  <bnoordhuis>`curl --head` says the files haven't changed recently though. maybe it's something that simply got overlooked in #586?
13:47:29  <italoacasas>Ok I know what is the problem
13:47:51  <bnoordhuis>(my comment was about latest-v0.12)
13:47:56  <italoacasas>Myles just confirm this with me in twitter
13:48:16  <jbergstroem>..and for 0.12.18, darwin, headers, and linux were created 8 hours post shasums creation
13:48:19  <italoacasas>Yesterday I promote by mistake 12.18 instead v7
13:48:26  <italoacasas>The first time
13:48:27  <bnoordhuis>ah
13:48:38  <jbergstroem>hangon
13:48:42  <jbergstroem>let me check if our backup still has old
13:48:54  <jbergstroem>(curse myself for not moving to incremental)
13:49:08  <italoacasas>I ctrl c before ends
13:49:51  <italoacasas>We review and everything was ok, but seems like nop...
13:49:53  <jbergstroem>yeees
13:49:56  <jbergstroem>i think we have the old stuff
13:50:28  <bnoordhuis>i'm kind of surprised the release job lets you overwrite existing files. is that intentional?
13:51:00  <italoacasas>Myles was thinking that was not possible
13:51:28  <bnoordhuis>seems we have evidence to the contrary :)
13:51:31  <jbergstroem>bnoordhuis: yes; we frequently run into issues wiht release jobs forcing us to rebuild
13:51:45  <jbergstroem>bnoordhuis: we should do a .DONE or similar though so after we deem it ok we lock it down
13:51:53  <bnoordhuis>good idea
13:53:21  <jbergstroem>so
13:53:46  <jbergstroem>do i copy over from backup ?
13:54:15  <bnoordhuis>if the checksums check out, go for it
13:54:32  <jbergstroem>its just an older version of hte folde that looks like it hasn't been modified the same way
13:54:37  <jbergstroem>let me back stuff up and we try, cool?
13:54:50  <bnoordhuis>yep
13:54:56  <jbergstroem>also: remind me to actually increment these things; we shouldn't generally screw up anyway
13:55:54  <bnoordhuis>"shouldn't screw up" is an admirable goal
13:57:13  <jbergstroem>done
13:57:47  <jbergstroem>bnoordhuis: what i meant was that hardlinks doesnt' grow in size -- just pointless _not_ to do incremental backups
13:58:06  <jbergstroem>anyone want to verify?
14:00:20  <jbergstroem>ah, looks like ben did
14:00:25  <bnoordhuis>yes, i did. nice work
14:01:40  <jbergstroem>i'll improve the backup routine hopefully later today
14:07:46  <italoacasas>thanks for resolving this so quickly... I will try to avoid promoting 12.18 again next week......
14:11:32  <bnoordhuis>it happens, no worries
14:14:09  <jbergstroem>italoacasas: no worries at all!
14:14:22  <jbergstroem>just good we managed to restore it
14:14:36  <italoacasas>yeap, thanks
14:43:55  <italoacasas>jbergstroem: are you around?
14:44:01  <jbergstroem>italoacasas: yes
14:44:12  <italoacasas>cool, question
14:44:38  <italoacasas>we release yesterday without armv6l done, then I'm promoting the build and updating the blog post
14:45:07  <italoacasas>I did the first but...
14:45:07  <italoacasas>I get this msg
14:45:09  <italoacasas>https://www.irccloud.com/pastebin/1Xw5uIV4/
14:45:37  <jbergstroem>is it a secondary key?
14:45:38  <jbergstroem>subkey, sorry
14:45:38  <italoacasas>but... in disk I can see arm6l
14:45:53  <italoacasas>dist*
14:45:53  <italoacasas>I just think is not signed
14:45:57  <italoacasas>is the correct key
14:48:32  <italoacasas>how I can check that the armv61 build is signed?
14:48:38  <jbergstroem>i'm just checking the keys
14:50:06  <italoacasas>thank you
14:50:26  <jbergstroem>the shasums for v6 is not in shasums.txt
14:50:30  <jbergstroem>so yeah they're missing
14:50:45  <italoacasas>any idea what happen?
14:50:55  <jbergstroem>looking into it
14:52:13  <italoacasas>well sorry, I know what happen, any idea **why**
14:53:07  <italoacasas>https://www.irccloud.com/pastebin/Wz8PA5Ot/
14:53:23  <italoacasas>this is the error
14:53:53  <jbergstroem>https://github.com/nodejs/node/blob/b315e2455e5121b6551c17bc7793ef0a101080f7/tools/release.sh#L72
14:53:57  <italoacasas>but, the build was promoted
14:54:25  <italoacasas>yes, Myles told me to comment the if statement, but when I try again the build was promoted already by me
14:54:30  <italoacasas>in the error try
14:54:57  <italoacasas>https://www.irccloud.com/pastebin/Ws24jKAF/
14:56:31  <italoacasas>O srry I see, first the script promote and then check the gpg key... That should be the other way around..
14:56:38  <jbergstroem>it lokos legit
14:56:46  <jbergstroem>*looks
14:57:02  <jbergstroem>i mean; the keys checks out
14:57:38  <jbergstroem>"using RSA key 23EFEFE93C4CFFFE" vs "3C4CFFFE"
14:57:44  <italoacasas>yeah I have only that one in this machine, I don't have more options to check
14:58:21  <jbergstroem>gpg --list-secret-keys | awk -F'( +|/)' '/^(sec|ssb)/{print $3}' gives you that key too?
14:58:45  <italoacasas>yeap
14:59:17  <italoacasas>is possible that the error is because I'm signing with my primary key?
14:59:34  <italoacasas>https://www.irccloud.com/pastebin/BFOTtCyx/
15:00:08  <jbergstroem>that's the key you should be using
15:00:28  <italoacasas>mm I was thinking that the second is the key people should use to sign
15:00:35  <italoacasas>nvm then
15:00:54  <italoacasas>ok
15:01:14  <jbergstroem>perhaps comment that line
15:01:19  <jbergstroem>if you can verify that its using that key
15:01:22  <jbergstroem>i mean it looks legit
15:01:31  <jbergstroem>i wonder what bails
15:01:37  <jbergstroem>do you get some other bash or something?
15:01:39  <italoacasas>what line?
15:01:58  <jbergstroem>72-75
15:02:09  <italoacasas>nop that is the full bash output
15:02:26  <italoacasas>yeah I did that already, but since the build is promoted already I don't have more the option
15:02:29  <jbergstroem>i mena; is /usr/bin/env bash a different bash version than the shell you're testing output from?
15:02:34  <jbergstroem>oh, right
15:03:16  <jbergstroem>perhaps grab the file and signature
15:03:17  <jbergstroem>and sing it
15:03:18  <italoacasas>nop about the bash question
15:03:18  <italoacasas>same
15:03:21  <jbergstroem>pass it to me
15:03:31  <jbergstroem>i mean; bake it manually
15:04:11  <italoacasas>can you walk me through the process?
15:04:33  <jbergstroem>https://github.com/nodejs/node/blob/b315e2455e5121b6551c17bc7793ef0a101080f7/tools/release.sh#L93
15:05:37  <jbergstroem>italoacasas: lets perhaps wait until myles or evanlucas is around; i don't want to mess with their established routines
15:05:48  <italoacasas>kk
15:06:01  <evanlucas>whats going on?
15:06:45  <italoacasas>do you have the IRC history available Evan?
15:06:49  <italoacasas>9:52
15:06:55  <evanlucas>yea looking
15:06:59  <jbergstroem>evanlucas: since armv6l was uploaded post shasum generation it is missing from shasums256.txt
15:07:18  <evanlucas>ah you should be able to resign the shasums256.txt using the release script
15:07:19  <jbergstroem>just wondering what your normal procedure for post-signing is
15:07:37  <evanlucas>I have always waited until all of the binaries have been built before promoting though
15:07:37  <jbergstroem>it (the release in question) doesn't seem to show up as an option
15:07:52  <evanlucas>at all?
15:07:52  <jbergstroem>i'm reading into dist-promotable
15:08:04  <evanlucas>what if you pass the -s v7.6.0 flag to the release script?
15:08:04  <jbergstroem>well, i'll let italoacasas take it from here but i'll look at the script
15:08:27  <italoacasas>thanks for everything jbergstroem
15:08:31  <jbergstroem>italoacasas: np
15:08:52  <italoacasas>evanlucas: the release was promoted and then I get the error
15:08:58  <italoacasas>I'm going to try the flag
15:09:31  <italoacasas>ok that works
15:10:58  <italoacasas>perfect, I can see the build here
15:10:59  <italoacasas>https://nodejs.org/dist/v7.6.0/SHASUMS256.txt
15:11:11  <italoacasas>thanks evanlucas
15:11:40  <evanlucas>np
16:18:16  * bnoordhuisquit (Ping timeout: 240 seconds)
16:36:26  * chorrellquit (Quit: My MacBook has gone to sleep. ZZZzzz…)
16:38:21  * bnoordhuisjoined
16:48:09  * bnoordhuisquit (Ping timeout: 260 seconds)
17:53:39  * bnoordhuisjoined
17:54:09  * not-an-aardvarkjoined
18:51:10  * chorrelljoined
19:46:34  * chorrellquit (Quit: Textual IRC Client: www.textualapp.com)
19:53:11  * lanceballchanged nick to lance|afk
20:49:46  * jenkins-monitorquit (Remote host closed the connection)
20:49:58  * jenkins-monitorjoined
21:43:10  * lance|afkchanged nick to lanceball
21:43:25  * node-ghjoined
21:43:25  * node-ghpart
21:45:02  * node-ghjoined
21:45:02  * node-ghpart
21:50:56  * node-ghjoined
21:50:56  * node-ghpart
22:10:16  * bnoordhuisquit (Quit: leaving)
23:41:35  <Trott>Oh, LinuxOne bot, I used to be so fond of you for being so reliable and running the tests faster than anything else in the farm. But now you've gone over to the dark side of failing on build lots. The love affair couldn't last forever.