[00:02:11] htop [01:10:14] 3operations, Wikimedia-Labs-wikitech-interface: fix wikitech-static updates - https://phabricator.wikimedia.org/T83596#1038380 (10Dzahn) [02:03:47] hoo: It looks up to me; do you still need a restart? [02:04:18] No, the tool owner replied to my mail by now (and I'm member of the project now) [02:04:23] thanks for your offer, though [02:04:56] Sorry it wasn't more timely; I'm outside my work hours. I keep an eye on IRC but lag varies. :-) [03:22:24] !dn labs is awesome [03:22:32] !dn act files $* into dev/null as requested. [03:22:33] Key was added [03:27:36] 3Wikimedia-Labs-General, Release-Engineering: Increase quota for deployment-prep (beta) project - https://phabricator.wikimedia.org/T73886#1038709 (10bd808) Current settings: Cores: 113/120 RAM: 231424/262144 Floating IPs: 7/7 Instances: 43/45 Security Groups: 8/10 #releng can decide if they... [03:54:43] !quota [03:54:43] virt1000 - nova quota-show --tenant | nova help quota-update [04:09:26] 3Tool-Labs: enwiki database dumps missing - https://phabricator.wikimedia.org/T89537#1038760 (10Betacommand) 3NEW a:3coren [05:42:17] Anyone around tonight who could help with Gerrit access? [05:44:08] 3Tool-Labs: Add wiki title case sensitivity flag (is_sensitive) to meta_p.wiki to support jbo.wp and wiktionary tools - https://phabricator.wikimedia.org/T69476#1038833 (10jayvdb) [05:44:17] I had to do a password reset in wikitech, but I still can't log in to Gerrit. Wondering if there is a lag in synchronizing. [06:16:48] 3Tool-Labs: Add wiki title case sensitivity flag (is_sensitive) to meta_p.wiki to support jbo.wp and wiktionary tools - https://phabricator.wikimedia.org/T69476#1038836 (10jayvdb) It would probably also be useful to add a jbo wiki to the wmfbeta farm, so testing of features can be conducted there with it's diffe... [11:26:19] 3Wikimedia-Labs-General, Tool-Labs: labswiki isn't replicated on Labs - https://phabricator.wikimedia.org/T89548#1038984 (10TTO) 3NEW [11:37:00] PROBLEM - Puppet staleness on tools-exec-15 is CRITICAL: CRITICAL: 100.00% of data above the critical threshold [43200.0] [18:19:04] PROBLEM - Free space - all mounts on tools-webgrid-02 is CRITICAL: CRITICAL: tools.tools-webgrid-02.diskspace.root.byte_percentfree.value (<62.50%) [20:45:50] andrewbogott_afk: if you are around and not afk in fact, short question please [20:47:15] Coren would work too [21:17:00] (03PS1) 10Florianschmidtwelzow: Add Gather to #wikimedia-mobile [labs/tools/grrrit] - 10https://gerrit.wikimedia.org/r/190677 [21:17:03] (03CR) 10jenkins-bot: [V: 04-1] Add Gather to #wikimedia-mobile [labs/tools/grrrit] - 10https://gerrit.wikimedia.org/r/190677 (owner: 10Florianschmidtwelzow) [21:17:45] (03PS2) 10Florianschmidtwelzow: Add Gather to #wikimedia-mobile [labs/tools/grrrit] - 10https://gerrit.wikimedia.org/r/190677 [22:16:32] matanya: It's not a typo. :-) [22:16:48] then what is it Coren ? :) [22:16:59] I replied in email. It's SGID [22:17:38] didn't get it, but does puppet know how to parse that ? [22:18:23] matanya: Puppet could not care less. That is simply an opaque value given to chmod() [22:18:35] I see [22:18:51] maybe a comment there would be a good idea [22:18:59] More to the point, those values are both correct and necessary. [22:19:26] ... why? That bit is no less normal a normal mode bit than the other eleven. [22:20:07] Coren: cause puppet-lint complians [22:20:17] file mode should be 4 digits [22:20:47] matanya: Then puppet-lint is broken. File modes are 12 bits and can therefore take five (octal) digits. [22:21:31] so what does 02775 mean in this case? [22:22:29] matanya: That the most significant three are most often zero is immaterial. Whoever wrote that puppet lint rule should "man 2 chmod" for sure. 02755 is ls -l "drwxrwsr-x" [22:22:39] (03CR) 10Merlijn van Deen: [C: 032] Add Gather to #wikimedia-mobile [labs/tools/grrrit] - 10https://gerrit.wikimedia.org/r/190677 (owner: 10Florianschmidtwelzow) [22:22:42] (03Merged) 10jenkins-bot: Add Gather to #wikimedia-mobile [labs/tools/grrrit] - 10https://gerrit.wikimedia.org/r/190677 (owner: 10Florianschmidtwelzow) [22:22:55] (Well, 'd' because it's a directory in this case - it could have been a file too) [22:23:20] Other example: temporary directories would normally be '01777' [22:23:23] Coren: https://docs.puppetlabs.com/references/latest/type.html#file-attribute-mode for the source of the confustion, i guess [22:24:04] Ah, indeed. Whoever wrote that bit of the puppet doc clearly doesn't actually understand how posix file modes actually work. :-) [22:24:37] OK, i'll open a bug report and a maybe a pull request [22:24:48] There is no such thing as "standard four-digit octal notation" beyond "most file modes, when represented in octal, happen to have four digits" [22:24:48] thanks for this! [22:24:48] !log tools.lolrrit-wm restarted for https://gerrit.wikimedia.org/r/190677 [22:24:51] Logged the message, Master [22:25:19] every day a new puppet issue uncovered :) [22:25:24] matanya: No worries. :-) [22:26:45] Alternately, those mode bits could be written out in long as 'u=rwx,g=rwxs,o=rx' but that's getting ridiculous. :-) [22:30:14] Coren: for ref: https://github.com/rodjek/puppet-lint/issues/394 [22:30:48] matanya: Ayup. [22:33:08] Coren: how would you say that part? I would like to push a PR for https://github.com/puppetlabs/puppet-docs/blob/master/source/references/3.7.4/type.markdown#file-attribute-mode [22:35:22] matanya: That paragraph is a little confused; it clearly knows there are twelve bits of mode that could be set, but has its example confuse the mandatory leading 0 for octal numbers with the fourth digit - or perhaps the leading 0 *isn't* mandatory in puppet as usual, in which case the example adds it for no good reason and fails to mention the deviation from convention. [22:35:47] yes, i agree [22:35:55] not written very well [22:36:05] (Most scripting languages demand a leading 0 to octal constants to separate from decimal) [22:36:27] Including ruby [22:37:22] And, for that matter, even if puppet actually accepted '2755' as an octal constant, that's be a really confusing thing and imo a bug. [22:38:58] Coren: line 69 in modules/git/manifests/clone.pp [22:40:18] Interesting. That means that the numeric literal accepted by puppet is forcibly in octal leading 0 or not? [22:40:29] * Coren boggles a little. [22:40:47] Coren: or that everyone is setting the wrong file modes =p [22:41:04] iirc the default is forcible 0 [22:41:11] ah, but those file modes are strings [22:41:15] "Sure, that convention has made numeric base clear since at least 1966 - let's ignore it" [22:41:30] where is that from ? [22:41:49] Coren: but it's not a /numeric literal/, it's a string composed of digits [22:41:56] matanya: My dramatization of what the puppet devs seem to have thought. [22:42:06] nice one [22:42:07] valhallasw`cloud: It's *used* as a numeric literal. [22:42:52] Coren: in what sense? everywhere in that puppet class it's a string [22:43:06] and as far as I know ruby does not silently coerce strings into integers [22:43:20] valhallasw`cloud: It's used as the numeric argument to chmod() [22:44:29] Ah, I was wrong - not 1966 but 1969. I thought BCPL used the leading zero convention but that was added in B three years later. [22:44:34] "The desired permissions mode for the file, in symbolic or numeric notation. This value should be specified as a quoted string; do not use un-quoted numbers to represent file modes." [22:44:45] Coren: chmod, the command line tool, /also/ doesn't require prepending zeros [22:45:37] valhallasw`cloud: That's... actually a reasonably sane argument for not requiring it. Hm. [22:46:52] Well, if nothing else it's a documentation bug. :-) [22:48:15] The documentation says it should a) be a string and b) either four octal digits or a comma-seperated list of permission clauses [22:49:06] I think I see what you mean. I suppose when, interpreted that way, it makes sense. [22:49:26] Coren: https://github.com/puppetlabs/puppet-docs/pull/453 [22:49:31] also, I'd argue the fact it's octal is somewhat irrelevant; people typically treat permissions as 'a magic digit no-one can remember what it does', 'owner permission digit', 'group digit', 'world digit'. The fact that those digits are actually octal values is somewhat irrelevant [22:49:41] Although the reference to "standard four-digit" is insane - there is no such standard. [22:50:56] valhallasw`cloud: ... then you'd argue in vain. Having documentation propagate misunderstanding is a bug regardless. :-) [22:51:05] Coren: would you report a bug against the fact puppet accepts 2755 ? [22:51:35] matanya: No, valhallasw`cloud made a reasonable case that it's not unreasonable to interpret the digit string as octal always. [22:51:47] Coren: it's the same thing as a phone number not actually being a number. It's a string composed of digits. [22:51:51] (In that the chmod command-line tool does the same) [22:52:07] valhallasw`cloud: Except that file modes very specifically is /not/ a string of digit but a bitfield. [22:52:23] valhallasw`cloud: That it can be *mis*interpreted that way is immaterial. [22:52:25] ok, so my work here is done :) a PR and a bug report. [22:52:32] maybe. [22:52:39] thanks a lot Coren (and valhallasw`cloud ofc) [22:55:02] valhallasw`cloud: (Incidentally, that is why I always tell people to use symbolic modes when using commandline chmod - they are less confusing for humans) [22:57:23] Coren: *nod*. I just have always trouble remembering what they are, while '755' and '644' are sort-of engrained in my memory :-p [22:58:18] for anything else, I have to replicate which bit meant which based on those two bitmasks [22:58:28] well, except putting zeros in. [22:59:20] Heh. I've been doing coding on unices since the late 80s; making the bits into an octal literal is reflex by now. :-) [23:33:54] 3Tool-Labs: Packages to be added to toollabs puppet - https://phabricator.wikimedia.org/T55704#1039604 (10scfc) [23:33:54] 3Tool-Labs: Install rake for Tools Labs - https://phabricator.wikimedia.org/T70208#1039602 (10scfc) 5stalled>3Invalid According to [[https://en.wikipedia.org/wiki/User:OverlordQ|his userpage]], the requester is on a long wikibreak. Please reopen this task when the need arises again.