| [11:04:59] | * theMusician has quit (Quit: theMusician) |
| [11:27:05] | * s-mutin has joined #aegir |
| [11:28:20] | * s-mutin has quit (Read error: Connection reset by peer) |
| [11:28:45] | * s-mutin has joined #aegir |
| [16:02:35] | * anarcat has quit (Ping timeout: 240 seconds) |
| [16:14:44] | * anarcat has joined #aegir |
| [17:03:33] | * ybabel has joined #aegir |
| [17:24:22] | * reaper013 has joined #aegir |
| [19:45:53] | * oluabbeys has joined #aegir |
| [22:26:05] | * oluabbeys1 has joined #aegir |
| [22:27:12] | * oluabbeys has quit (Ping timeout: 260 seconds) |
| [22:27:12] | * oluabbeys1 is now known as oluabbeys |
| [22:40:05] | * oluabbeys has quit (Ping timeout: 240 seconds) |
| [23:10:19] | <ergonlogic[m]> | hi all, sorry for disappearing. it's been a crazy summer |
| [23:12:18] | <ergonlogic[m]> | hefring: seen memtkmcc? |
| [23:12:18] | <hefring> | memtkmcc was last seen in #aegir 4 months 3 weeks ago saying 'OK, Aegir works with ProxySQL, but the trick was to force Provision to perform all tests and backend tasks by talking to writer db node directly (without ProxySQL), plus on all Galera nodes you need to enable pxc_strict_mode=PERMISSIVE, so it will not break things due to tables without an explicit primary key — but this should be fixed in Aegir, and is easy to spot in the syslog where Galera tells you whic'. |
| [23:13:17] | <ergonlogic[m]> | anybody running webpack on nginx? |
| [23:16:18] | * oluabbeys has joined #aegir |
| [23:36:33] | * oluabbeys has quit (Ping timeout: 248 seconds) |
| [23:43:29] | <helmo42[m]> | ergonlogic[m]: not me |
| [23:43:37] | <helmo42[m]> | but good to see you here :) |
| [23:44:14] | <helmo42[m]> | Summer got in the way for me too ... we're overdue for a release... What about next week? |
| [23:50:20] | <helmo42[m]> | Please post any release blockers on https://www.drupal.org/node/2882434 |
| [23:50:21] | <hefring> | https://www.drupal.org/node/2882434 => [meta] 3.12 release (bugfix/patches) [#2882434] => 5 comments, 1 IRC mention |
| [23:55:02] | <ergonlogic[m]> | helmo42: has memtkmcc been around? |
| [00:07:28] | <ergonlogic[m]> | I think the nginx/web_pack issue I was having was just that the site's settings.php wasn't being chgrp'd www-data... |
| [00:09:54] | <ergonlogic[m]> | probably mis-configured nfs... |
| [00:20:02] | * theMusician has joined #aegir |
| [00:29:55] | <ergonlogic[m]> | nope... it looks like the master server's being chgrp'd to 'aegir'... |
| [00:53:55] | <helmo42[m]> | ergonlogic[m]: It's been quiet here during the summer |
| [00:59:05] | * theMusician has quit (Ping timeout: 240 seconds) |
| [01:00:23] | * reaper013 has quit (Quit: Page closed) |
| [01:23:27] | * hefring has joined #aegir |
| [01:31:58] | <ergonlogic[m]> | hmm, looks like it may just be that I hadn't added the aegir user to the www-data group on the cluster servers... |
| [01:52:12] | * mstenta has joined #aegir |
| [02:21:08] | * ybabel has quit (Quit: ybabel) |
| [02:21:26] | * ybabel has joined #aegir |
| [02:35:39] | * shaneonabike1 has joined #aegir |
| [02:37:40] | <jonpugh[m]> | ergonlogic: I was just testing out fix_permissions and ownership the other day... is there a reason not to include them in Aegir core? |
| [02:40:05] | * theMusician has joined #aegir |
| [02:57:51] | <helmo42[m]> | ergonlogic[m]: they are in tasks_extra which is in core ??? right? |
| [03:02:21] | <ergonlogic[m]> | I believe so |
| [04:04:39] | * shaneonabike1 has quit (Quit: Leaving.) |
| [04:21:37] | * shaneonabike1 has joined #aegir |
| [04:28:01] | <jonpugh[m]> | I don't mean in a contrib module for "extra tasks", I mean, provision itself should use those permissions and ownership rules and scripts, shouldn't it? |
| [04:28:20] | <jonpugh[m]> | that module enforces different permissions than provision itself. |
| [04:28:48] | <jonpugh[m]> | helmo42: ergonlogic: I suppose I meant "a part of provision" not just "included in aegir core" |
| [04:47:16] | * shaneonabike2 has joined #aegir |
| [04:47:54] | <ergonlogic[m]> | it's been a while since I wrote those scripts, but the intention was to mirror the ownership and permissions in Provision. The main thing is that they're run under sudo, and so can fix stuff that provision can;t on it's own |
| [04:48:29] | <jonpugh[m]> | yep |
| [04:48:35] | * shaneonabike1 has quit (Ping timeout: 240 seconds) |
| [04:48:41] | <jonpugh[m]> | I just noticed that they don't mirror those permissions |
| [04:49:15] | <jonpugh[m]> | and since that's kindof the point of provision and provision-verify, shouldn't that functionality just be a part of provision itself? |
| [04:49:53] | <jonpugh[m]> | most users assume that's what is happening on provision-verify anyway, myself included |
| [04:53:45] | <ergonlogic[m]> | part of the challenge with integrating this more with Provision, is that it can't safely be implemented in a Drush command, nor should it pull a list of dirs to chmod/chown from Drush, because Drush is so promiscuous about loading commandfiles. |
| [04:54:18] | <ergonlogic[m]> | I'm open to suggestions, but it's kinda playing with fire |
| [04:54:42] | <jonpugh[m]> | but isn't that what the module does? |
| [04:54:44] | * jonpugh[m] sent a long message: jonpugh[m]_2017-09-06_18:54:43.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/PIPqwnsBXPwEdvLW... |
| [04:55:04] | <jonpugh[m]> | the drush command just runs drush_shell_exec |
| [04:55:40] | <ergonlogic[m]> | yes, but the script validates (poorly) that it's running in a platform or site context |
| [04:56:46] | <ergonlogic[m]> | so that you can't chown:chmod individual files or arbitrary directories. /etc/groups for example |
| [04:57:14] | <jonpugh[m]> | I see... |
| [04:57:16] | * jonpugh[m] sent a long message: jonpugh[m]_2017-09-06_18:57:16.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/NCiKDvuRrgqrPkwk... |
| [04:57:34] | <ergonlogic[m]> | right |
| [04:57:55] | <jonpugh[m]> | perhaps the shell script can switch back to aegir user to run a `drush @site status` or something |
| [04:59:49] | <jonpugh[m]> | wait a minute, shouldn't that "||" be an "&&"? the rest get skipped if the drupal_root directory exists at all? |
| [05:02:18] | <jonpugh[m]> | sorry nevermind. forgot what -z is for |
| [05:20:00] | <ergonlogic[m]> | feel free to add some docs if anything is unclear |
| [05:40:10] | * ybabel has quit (Quit: ybabel) |
| [07:03:13] | * mutin-s has joined #aegir |
| [07:06:39] | * s-mutin has quit (Ping timeout: 248 seconds) |
