[00:00:21] (Merged) jenkins-bot: Split error check into own function [wikimedia/fundraising/SmashPig] - https://gerrit.wikimedia.org/r/340905 (owner: Ejegg) [00:12:45] hmm, updb claims to still be running, but processlist shows no evidence of the geo queries [00:17:29] d'oh. [00:17:46] you're running show processlist as the civicrm service user? [00:17:50] ejegg: ^ [00:20:36] PayPal "English Cuisine" [00:23:03] awight: yep, under drush [00:24:31] pstree maybe? [00:29:59] oh nifty [00:31:04] and, looks like it finished [00:31:21] but my update threw an error whose details are not shown [00:32:48] Aaargh. [00:32:53] I always drush -v -v $* [00:33:13] cwd|afk: ^ would you please add that to the drush wrapper? [00:33:20] yep, I -v -v 'ed it [00:33:32] awnuts [00:33:33] but it's spitting out a lame HTML error [00:33:46] oh hey, syntax error. [00:33:50] dangit [00:33:52] * awight changes hue [00:33:59] ah hehe n.b.d. [00:34:01] thought i ran that on my machine verbatim [00:35:49] d'oh, commentfail [00:36:27] -- not at beginning of line [00:37:51] cwd|afk: we could also use a "-r {::crm::root_dir}/drupal" so it can be run from anywhere--if you think that's prudent? [00:38:40] (PS1) Ejegg: Fix busted SQL comment [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340908 [00:39:00] ejegg: You ran Eileen's updb? [00:39:31] * awight gives the server admin log a pointed look ;-) [00:40:21] awight should be there, unless stashbot's busted! [00:41:04] Fundraising Sprint Baudelaire Bowdlerizer, Fundraising Sprint Costlier Alternative, Fundraising Sprint Deferential Equations, Fundraising Sprint English Cuisine, and 4 others: Purge Varnish cache when a banner is saved - https://phabricator.wikimedia.org/T154954#3069707 (awight) [00:41:23] !log CiviCRM geocoding update finished, name fix failed on badly formatted comment [00:41:27] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [00:42:00] !log re-enabled CiviCRM de-dupe jobs [00:42:04] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [00:42:18] awight comment fix: https://gerrit.wikimedia.org/r/340908 [00:51:02] mind giving that trivial thing a +2 ^^^ ? [00:51:09] sorry, gotcha 23:49 ejegg: running batched geocoding update and donor name fixes [00:51:12] yep [00:51:25] ty! [00:51:44] (CR) Awight: [C: 2] Fix busted SQL comment [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340908 (owner: Ejegg) [00:56:32] hi eileen1 ! turned out there was an easy way to get the names back from the db, no log parsing required [00:57:02] also, I just ran your geocoding redo and empty address table script [00:57:21] (Merged) jenkins-bot: Fix busted SQL comment [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340908 (owner: Ejegg) [00:59:42] ejegg: great + phew! [00:59:57] or the reverse orer perhaps [01:00:21] (PS1) Ejegg: Merge branch 'master' into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340910 [01:00:31] (CR) Ejegg: [C: 2] Merge branch 'master' into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340910 (owner: Ejegg) [01:00:37] (Merged) jenkins-bot: Merge branch 'master' into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340910 (owner: Ejegg) [01:02:21] hehe, !log updated civicrm from 2d1de872d5b3ef81a1f2918e73dd7088af9844ed to 0cab1930bbc0a91c435caef4e8a80da24b514f90 [01:02:37] !log re-running fix for missing names [01:02:42] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [01:03:34] * ejegg facepalm [01:03:39]
DB Error: unknown error
[01:03:42] grr [01:06:20] works locally, verbatim [01:08:09] weird, i did exactly zero things in riot in the last 18 hours [01:18:19] (PS1) Ejegg: Just zap the comment [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340912 [01:18:46] fr-tech my latest flailing attempt to make that run on prod ^^^ [01:18:53] crud, almost out of battery [01:20:43] wait, I should be flailing on staging [01:21:06] hmm, not that there are any rows to update there [01:22:31] (CR) Awight: [C: 2] "tersely put :)" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340912 (owner: Ejegg) [01:22:39] ejegg: Want me to finish this up? [01:23:00] awight, please do if you've got the time! [01:23:07] Been sitting on my hands and barely reviewing things, so I have an excuse to work late, heck yeah! [01:23:24] I think you handed me the winning script contents anyway, so it'll be a breeze [01:23:32] worst case is another nightly silverpop export with bad data [01:23:34] Enjoy the summer night! [01:23:37] thanks! [01:23:51] * awight pulls hoodie tight [01:26:20] (Merged) jenkins-bot: Just zap the comment [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340912 (owner: Ejegg) [01:26:40] k, gnight all [01:27:08] (PS1) Awight: Just zap the comment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340913 [01:27:26] (CR) Awight: [C: 2] Just zap the comment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340913 (owner: Awight) [01:27:33] (Merged) jenkins-bot: Just zap the comment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340913 (owner: Awight) [01:28:02] !log update civicrm from 0cab1930bbc0a91c435caef4e8a80da24b514f90 to 133bde2cbffd8e5da01c64a8b27c241d8f089c96 [01:28:07] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [01:29:11]
DB Error: unknown error
[01:35:01] It must have failed at the second statement [01:35:25] cos the 60 rows that haven't been converted all have the email in sort_name [01:35:46] Luckily, the first statement is safe to rerun. [01:45:32] !log rerun schema change wmf_civicrm:7480 [01:45:36] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [01:45:38] * awight rolls the dice [01:49:34] Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM: CiviCRM drush error message should never be HTML-formatted. And should prominently contain useful debugging info. - https://phabricator.wikimedia.org/T159506#3069802 (awight) [02:07:02] (PS1) Awight: Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) [02:07:13] eileen1: ^ Whenever you have a moment [02:07:40] It turned out, there was something trigger-related which caused the staging query to succeed and the production one to explode. [02:10:51] awight: looks ok - did you try running it on staging or anywhere? [02:11:42] actually - aren't you running two queries together? [02:12:07] Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM, fundraising-tech-ops: CiviCRM triggers should be the same on staging as on production - https://phabricator.wikimedia.org/T159507#3069824 (awight) [02:12:18] Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM, fundraising-tech-ops: CiviCRM triggers should be the same on staging as on production - https://phabricator.wikimedia.org/T159507#3069839 (awight) [02:12:25] (CR) Eileen: Use an intermediate temporary table so we don't step on a trigger landmine (1 comment) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) (owner: Awight) [02:12:46] eileen1: yah I ran that query with just the table names changed [02:13:00] of course, the original problem is something that only appears on production :) [02:13:12] though I don't think this new query is fragile in that same way [02:13:20] eileen1: ooh [02:13:24] thx now I see what you mean [02:13:33] yeah I can't run two queries thus! [02:13:49] so, just to understand - why do we need to involve the log table - isn't the data in sort_name in the civicrm_contact table? [02:14:37] (PS2) Awight: Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) [02:14:57] eileen1: These are the corner cases in which we overwrote the sort_name with an email, for unknown reasons. [02:15:03] That's probably its own bug to investigate later. [02:15:29] There are 83 or so of these, as opposed to 6,800 of the ones where sort_name was recover, able [02:15:29] ah ok [02:15:57] so, when joining on the log_civicrm_contact table, there are no instances of more than one record in that table? [02:16:00] And, thanks for having the sun at a cooky angle down there :) [02:16:24] eileen1: oh good point, I was asking ejegg about that wrt the query I'm replacing. [02:16:52] There might still not be more than one record, since the update failed. [02:17:09] If there is a duplicate, I don't understand what would happen. [02:17:21] err s/duplicate/second audit log record for the contact [02:17:31] yeah [02:17:44] what about doing the update from the contact table as one run [02:17:49] & then the log table second [02:17:57] that way we'll get most from the contact table [02:18:06] & only the 83 will have the extra element of rist [02:18:08] risk [02:19:02] oh excellent idea, one moment pls [02:26:12] (CR) Eileen: Use an intermediate temporary table so we don't step on a trigger landmine (1 comment) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) (owner: Awight) [02:29:06] (PS3) Awight: Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) [02:29:36] eileen1: I might have found another problem with this backfill, that we aren't refreshing the sort_name and display_name. [02:29:58] If I set them to null will CiviCRM recalculate? [02:30:26] I suspect that if you set first_name & last_name through the api civi will calculate sort_name & display_name [02:32:04] Cool, that'll be a good refresher in how the Civi API :) [02:39:01] awight I took a quick check but we might have hacked something out - should we get something through now - or leave til next week & work through more slowly? [02:39:50] (PS4) Awight: Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) [02:39:56] Smoke testing locally... [02:40:40] I'm happy to hold off if anything looks fishy, the remaining data corruption has a pretty small impact at this point. [02:42:46] so, why and sort_name not like '%, %' [02:43:16] do they have blank sort_names too? I guess I missed that [02:43:49] copypasta... [02:43:52] lemme catch up with you [02:44:09] nah I think I've caught up with you now [02:46:18] (CR) Awight: Use an intermediate temporary table so we don't step on a trigger landmine (1 comment) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) (owner: Awight) [02:47:54] awight: btw civicrm_api3 throws exceptions so if ($result['is_error']) { will never happen [02:47:54] (PS5) Awight: Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) [02:48:06] (which is not a biggie here) [02:48:58] oops! I thought I was copying from recent Civi code [02:49:31] Great to hear it though, I think the "throw" is a much more usable function signature [02:50:02] (PS6) Awight: Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) [02:50:12] yeah I like exceptions [02:50:31] if you use civicrm_api() it returns an error but civicrm_api3() throws [02:50:42] I'm not great at choosing which level to catch exceptions on [02:50:56] ... wish I had more mentorship and stuff about that [02:51:56] It ran cleanly on staging, checking the results now. [02:53:46] I don't think it fixed any records. [02:55:22] honestly - I'm inclined NOT to catch exceptions if there is no known reason for there to be an error - in a one-off script anyway [02:55:36] if it needs to fail it should fail :-) [02:56:01] I don't think it fixed any records. From sort_name POV or first_name? [02:56:04] Yeah, fail fast is the best [02:56:27] "POV"? [02:56:52] The count for first_name=null and last_name=null didn't change after running the update... [02:57:55] ok [02:58:08] * awight facepalms for succumbing to the "temporary" table [02:58:08] but, is staging data up to date? [02:58:22] good question! d'oh [02:58:58] | 2017-02-02 09:42:02 | [02:59:28] cwd|afk: ^ something we'd like to automate and be able to kick of as techsen. [02:59:46] * awight burns incense and calf grease [02:59:56] awight: I'm going to read one more time & if it looks safe will approve & you can run on live [03:00:02] :) [03:00:52] wikibugs_: help [03:00:59] wikibugs_: why don't you talk to us any more [03:03:00] (CR) Eileen: [C: 2] Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) (owner: Awight) [03:03:17] ok - I can't see any room for it to do anything odd [03:04:32] (PS1) Awight: [WIP] Random fix to deprecated Civi code [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/340917 [03:05:28] * awight goes to plug patch into a 220V outlet [03:06:28] brb, I need to hide the bird food from rats [03:06:52] (Merged) jenkins-bot: Use an intermediate temporary table so we don't step on a trigger landmine [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/340914 (https://phabricator.wikimedia.org/T159455) (owner: Awight) [03:09:50] (CR) Eileen: "True - those lines will never be hit. We should open a JIRA & add a pr for this." [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/340917 (owner: Awight) [03:11:49] thanks for the note--I should do that to stay familiar with the process... [03:11:55] Seems a little heavyweight! [03:12:27] I guess that would encourage me to find a round dozen occurrences. [03:13:37] well upstreaming can be a bit of a drag [03:14:34] I don't know whether to encourage you to do it or not [03:16:46] :( [03:16:58] That's sort of the position I've heard people articulate about editing the wiki. [03:17:23] We could do things to make it more popular and visible, but nobody feels quite right about throwing new people under our multi-wheeled passenger vehicle [03:18:06] o/ [03:18:15] \o [03:19:09] cwd: have you had a chance to check out any of those radio stations? [03:19:14] KALX is kicking the butt ATM [03:19:33] awight: i have perused them all yes [03:19:40] internet radio is where it is at [03:20:02] I want it in my hoopty cars [03:21:08] awight: upstreaming to Civi is becoming harder not easier - but that's partly because more QA is going into every patch & the amount of qa resource has not grown [03:21:18] * awight marvels at how crApple is considered design king as I pick staples and other detritus out of supermagnetic adapter port [03:21:35] eileen1: That's sort of mixed promising :) [03:21:42] :-) [03:22:08] this is one of my favorite stations of all time, out of boulder http://radio1190.colorado.edu:8000/master.mp3 [03:22:19] TY [03:23:26] cwd: Let's figure out what format "mix" DJ/station playlists should be in. It's not quite m3u. [03:24:11] hmm yeah good question [03:24:15] awight: vimwiki! [03:24:59] Well, maybe it is m3u, but the DJ schedule is included as metadata [03:25:22] (PS1) Awight: Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340918 [03:25:34] (CR) Awight: [C: 2] Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340918 (owner: Awight) [03:25:40] (Merged) jenkins-bot: Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/340918 (owner: Awight) [03:26:29] !log update civicrm from 133bde2cbffd8e5da01c64a8b27c241d8f089c96 to d20ed40b52180252a99ecbcb4bb24308e10bc1a4 [03:26:34] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [03:27:16] !log rerunning schema_update wmf_civicrm:7480 [03:27:20] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [03:27:41] success... checking the data. [03:28:59] Looks good! [03:30:19] eileen1: Do you agree that issue (2) in https://phabricator.wikimedia.org/T159455 is probably unrelated? [03:30:25] awight: making notes of your requests from earlier, not ignoring! [03:31:34] eileen1: Argh, there are three issues in that bug! [03:32:54] cwd: /me burns more organs [03:33:57] awight: i guess i didn't know we were using a drush wrapper :S [03:34:12] Pretty smelly, huh? [03:34:26] It was the only thing we could think of to solve the file permission problems forever. [03:34:47] I thought that umask 0002 would be helpful in general, but not sure we've gone there yet. [03:35:25] Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM: Something is corrupting contact sort_name by copying in the email address - https://phabricator.wikimedia.org/T159508#3069881 (awight) [03:38:29] eileen1: O_o I don't understand this. I just checkedout civicrm-core#master and that EmailProcessor call uses civicrm_api rather than civicrm_api3 [03:38:32] reading the git logs... [03:38:53] it might be cos the calling code doesn't handle exceptions... [03:39:59] awight: sorry i'm rusty/ignorant, 002 would make files drush creates be RO for regular users? [03:40:47] awight: we are about 6 weeks behind master - maybe it has changed [03:40:59] The usual umask is 022, so 002 makes g+w the default [03:41:07] sorry, I'm confused - what needs solving now [03:41:18] I thought the EmailProcessor thing was a tangent [03:41:24] oh gotcha [03:41:51] eileen1: lol I'm not going crazy--you did a custom patch for WMF in "Let's create some noise!" which included the civicrm_api+3+ change [03:41:58] totally tangential [03:42:11] and the commit summary make me happy enuf to drop the trail [03:42:15] ah yeah - I remember that - I was debugging something [03:42:25] & I fixed something or other upstream [03:42:39] I was only concerned cos my conception of time only going forward was thrown off for a moment. [03:42:58] I like to think time only goes forward.... [03:44:26] Well, the head commit on upstream master was authored by you, and from Mar 3. Our branch was on an older commit but had newer code, also last changed by you. [03:44:48] crazy kids. [03:45:19] it is mar 3 here - maybe not there? [03:45:38] only on the east coast! [03:45:49] mostly I do that stuff to mess with your head, occassionally it has useful functionality too [03:46:13] lmao [03:46:45] the commit summary "let's create some noise" was an extra excellent touch [03:47:25] I had to look up lmao - I was thinking 'let me outta here' but that would be lmoh [03:47:39] hehe [03:48:04] ha [03:48:30] can you believe some civi marketting folks tried to convince me to stop writing some of the things I add into the code because they thought it was a bad look [03:48:53] I mean - if your purchasers are looking at the code.... [03:49:36] The one they complained about was return $this->hackyHandleBrokenCode(self::$jobLog); [03:49:48] but, I stand by it as an accurate function name [03:51:18] Fundraising Sprint Baudelaire Bowdlerizer, Fundraising Sprint Deferential Equations, Fundraising Sprint English Cuisine, Fundraising-Backlog: fill out PCI SAQ-A form for 2017 - https://phabricator.wikimedia.org/T155779#2954361 (awight) What I think when I hear "your compliance expires soon": {F60... [03:51:37] hehe eileen1 that's truly "open source" [03:51:41] awight: so, is that the end of the fall out from my mistake? [03:52:30] cwd you think I would have been fine if it were closed source & compiled to an exe [03:53:34] Fundraising Sprint Baudelaire Bowdlerizer, Fundraising Sprint Deferential Equations, Fundraising Sprint English Cuisine, Fundraising-Backlog: fill out PCI SAQ-A form for 2017 - https://phabricator.wikimedia.org/T155779#3069958 (Ejegg) Is that from the Omni Consumer Products robocop division? [03:54:20] eileen1: yes--but oddly enough there's no task to close ;) so it doesn't count against your pay [03:55:16] fantastic - I do feel bad tho [03:56:57] awight: otoh the log tables proved themselves [03:57:01] YES [03:57:09] I was thinking that too. One of many proofs [03:58:30] cwd if you are still around - it would be fantastic to get a staging restore kicked off tomorrow so I can test my address updates next week - doesn't need to be latest code [03:59:06] eileen1: i was just looking at that [03:59:43] i pushed the db export over earlier, gotta check if i have icinga access to silence the replag alerts [04:00:06] agreed with awight|familias earlier that we should automate this [04:00:30] code situation is perfect, IMO. We can use git or directly hack [04:04:22] yup agreed [04:05:23] only hard part as i can tell is 11GB db dump [04:05:36] necessary replag [04:05:44] yeah - it is also good not to keep it too up to date sometimes [04:05:57] ie. if you want to check if recent updates have changed things [04:06:21] mmm good point [04:06:26] (you can check from the logs but visual is nice) [04:07:12] but it should be not an su action [04:21:17] ejegg|away: Hey, caught you working. [04:21:25] ejegg|away: can you point me to the SA [04:21:33] SAQ thing for review? [04:21:53] nvm, got it. [04:21:57] ejegg|away: nvm! found it [04:36:53] Fundraising Sprint Baudelaire Bowdlerizer, Fundraising Sprint Deferential Equations, Fundraising Sprint English Cuisine, Fundraising-Backlog: fill out PCI SAQ-A form for 2017 - https://phabricator.wikimedia.org/T155779#3069981 (awight) :D +1'd the email "2017 PCI SAQ-A for review", I think this... [04:37:01] eileen1: just civi restore and no drupal right? [04:38:25] (CR) Awight: [C: 2] Use HashBag instead of EmptyBag [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/339330 (https://phabricator.wikimedia.org/T128692) (owner: Ejegg) [04:39:35] Fundraising Sprint Deferential Equations, Fundraising Sprint English Cuisine, Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM, Patch-For-Review: We still have contacts whose merge is blocked on mis-rounded geocodes - https://phabricator.wikimedia.org/T158271#3031937 (awight) Looks like thi... [04:40:22] cwd yep [04:40:43] cwd tbh I could live with just address & contact table - but not sure it's any easier [04:40:47] (Merged) jenkins-bot: Use HashBag instead of EmptyBag [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/339330 (https://phabricator.wikimedia.org/T128692) (owner: Ejegg) [04:41:15] awight|familias: I missed the meeting where we agreed to all carry hash-bags - isn't that illegal in CA? [04:42:04] Fundraising Sprint Deferential Equations, Fundraising Sprint English Cuisine, Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM, Patch-For-Review: We still have contacts whose merge is blocked on mis-rounded geocodes - https://phabricator.wikimedia.org/T158271#3069994 (Eileenmcnaughton) yeah... [04:42:30] eileen1: would definitely be faster except i'd have to tease them from the giant db dump [04:42:43] Fundraising Sprint Autotune Earphones, Fundraising Sprint Baudelaire Bowdlerizer, Fundraising Sprint Costlier Alternative, Fundraising Sprint Deferential Equations, and 6 others: [Spike] investigate contribution tracking data (was Engage import fail... - https://phabricator.wikimedia.org/T146295#3069995 [04:43:02] O_o [04:43:12] I should visit this chat room more often [04:43:55] jeff wrote a script to remove the trigger blocks which are incompat with the staging daemon, for some reason [04:43:56] cwd that was wrt the db restore or the hash bag? [04:43:58] eileen1: I didn't even catch the trollish commit summary [04:44:08] It's just normal here [04:44:12] eileen1: heh sorry the db restore [04:44:25] Pretty funny transferring between the Richmond and Pittsburg subways... [04:44:33] :-) [04:44:38] The Richmond one goes through Berkeley and always has a strong mix of odeurs [04:44:51] Pittsburg is into the deep 'burbs. [04:45:20] heh that subway route sounds decidedly east coast [04:45:52] or at least eastern 1/3 of US [04:46:11] It's one of those silly arrangements where the name of the line is the last stop, and the board of directors is obsessively extending the line every year in order to charge us more and make their buddies' real estate profitable. [04:46:21] So stay tuned for next year's line-up [04:46:34] hehe yeah, it's no Springfield [04:46:47] heh, it'll be practically philly [04:48:10] Your visiting friend go to sleep? Internet too plentiful now? [04:48:20] * awight looks at timezones suspiciously [04:48:42] gotta use up the MBs! [04:53:35] (PS1) Awight: Remove deprecated key-value interfaces [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/340924 (https://phabricator.wikimedia.org/T159175) [04:54:29] That's all I got. [04:54:39] See you tomorrow [04:55:59] (CR) jerkins-bot: [V: -1] Remove deprecated key-value interfaces [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/340924 (https://phabricator.wikimedia.org/T159175) (owner: Awight) [05:02:00] eileen1: sorry, i need to leave this db restore for the morning, there are some involved steps i intend to automate but it will take some fiddling. ok if it waits till the (US) morning? [05:04:07] cwd - I wanted to try to get it kicked off this week so it would be available next week [05:04:28] so sometime tomorrow would be fab [05:04:49] eileen1: i can definitely get it done tomorrow! [05:05:09] I would be very grateful [05:05:30] I will back up the address table & log_address table before testing this time [05:06:29] eileen1: no problem at all! bed time here, have a good night/evening! [16:07:44] argh that ghost matrix account is back and stealing my nick every night again [17:15:45] cwd: I'm enjoying that sentence ^^^ [17:15:54] but sorry about the nick stealing [17:16:46] i kicked it from every room but it still connects to the server at seemingly random intervals [17:17:58] having fun imagining what sort of brain damaged accidental AI the irc bridge spawned [17:44:00] argh, paypal audit parser is still getting errors looking up details [18:00:21] fr-tech: BE ALERT!!!! (The world needs more lerts...) [18:00:21] -- discuss. [18:16:07] fr-tech I'll hop in the talk in 2 minutes! [18:16:56] i am trying to resolve some freenode weirdness atm [18:17:33] k [18:18:00] I'm also around if it's useful... [18:31:04] so... one task I've been procrastinating hard on is the data retention. Want to chat about that? [18:31:07] AndyRussG: ^^^ [18:31:11] I'm in the hangout now [18:46:05] ejegg: coming! [18:46:07] restoring civi staging db [18:48:45] cwd ah, cool! [18:49:06] also looking at how it could be automated...there are a lot of steps [18:49:15] i was tempted to do a friday deploy of that iDEAL stuff, but I've come to my senses [18:53:23] people care about iDEAL [19:09:22] ejegg: so the thing about letting fr-tech fire off staging db updates is it takes 12 hours to import [19:09:27] which really limits the usefulness [19:12:43] it's totally possible, but probably a lot of work for little reward [19:12:58] if we can figure out a way to make it faster, it would be a lot more worthwhile [19:17:31] cwd my thought was that we'd kick it off before signing off for the night [19:18:39] Partly to avoid bugging Jeff_Green late in his day. But now we've got an op in a later time zone, so maybe it's not so necessary [19:19:20] yeah, well i documented what i did this time so let's see what actually requires su... [19:19:38] really just the mysql command [19:20:10] so there's an issue with importing the dump so it has to get some stuff stripped out first [19:20:12] the triggers [19:20:17] then inserted those separately [19:20:40] and it's 178G unpacked so it's easy to accidentally the whole disk [19:33:03] whew! [19:33:10] ginormous [19:33:37] i think there are some huge tables we never do anything with. [19:33:52] would be cool to get a size breakdown and see if there's stuff we can just truncate [19:33:58] yeah and we probably dump and restore 99.9% of the same data every time [19:34:31] it's a hard problem [19:35:34] the civi db is really complicated with all kinds of data structures inside it so taking just a slice is not reliable [19:46:08] I think I remember a log table we bolted on that's huge and unuseful. [19:50:17] it would certainly help to reclaim some space [19:50:26] but probably good to just deal with it being huge too [20:06:44] brb, buying a machete :) [21:23:37] Fundraising-Backlog, fundraising-tech-ops: Make it possible for fr-tech to acknowledge icinga alerts - https://phabricator.wikimedia.org/T152725#3072060 (Jgreen) [21:45:03] Fundraising-Backlog, fundraising-tech-ops: Make it possible for fr-tech to acknowledge icinga alerts - https://phabricator.wikimedia.org/T152725#3072117 (RobH) [21:45:06] fundraising-tech-ops, Operations: adding fundraising to icinga to ack alerts - https://phabricator.wikimedia.org/T159576#3072114 (RobH) Open>Resolved [21:45:49] Fundraising-Backlog, fundraising-tech-ops: Make it possible for fr-tech to acknowledge icinga alerts - https://phabricator.wikimedia.org/T152725#3072118 (Jgreen) This ~should~ be fixed as of today. Members of fr-tech should be able to ack alerts for any fundraising host/service they're alerted about, usi... [21:46:07] Fundraising-Backlog, fundraising-tech-ops: Make it possible for fr-tech to acknowledge icinga alerts - https://phabricator.wikimedia.org/T152725#3072119 (Jgreen) Open>Resolved [21:51:42] Fundraising-Backlog, fundraising-tech-ops: Make it possible for fr-tech to acknowledge icinga alerts - https://phabricator.wikimedia.org/T152725#2858158 (Ejegg) Rockin, thanks! Will this take a little while to go into effect? I just tried logging in with my LDAP username and password at https://icinga.wi... [22:07:59] XenoRyet: i dumped some normalized recurring messages here in case they might help write tests: https://gerrit.wikimedia.org/r/338145 [22:09:43] sweet, thanks [22:10:26] Oh, yea, I had seen those earlier. Helpful! [23:38:22] Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM: Something is corrupting contact sort_name by copying in the email address - https://phabricator.wikimedia.org/T159508#3072401 (Ejegg) Open>Invalid This only happens when first_name and last_name are null. Pretty sure Civi treats sort & display n... [23:59:14] ejegg: how hard will this be? https://phabricator.wikimedia.org/T159385