IRC logs for #aegir, 2016-12-30 (GMT)

2016-12-29
2016-12-31
TimeNickMessage
[11:20:45]* theMusician has quit (Quit: theMusician)
[17:14:02]* boshtian has joined #aegir
[20:10:07]* ybabel has joined #aegir
[22:54:36]* mstenta has quit (Quit: Leaving)
[23:07:27]* mstenta has joined #aegir
[00:37:09]* boshtian has quit (Ping timeout: 248 seconds)
[01:07:11]* boshtian has joined #aegir
[01:37:36]* fatguylaughing has joined #aegir
[03:36:07]* boshtian has quit (Quit: boshtian)
[05:27:16]* ybabel has quit (Quit: ybabel)
[05:57:34]* memtkmcc has joined #aegir
[05:59:18]<memtkmcc>hefring: tell colan: https://www.drupal.org/node/2839973#comment-11846690
[05:59:19]<hefring>https://www.drupal.org/node/2839973 => Rebuild the registry before database updates are run [#2839973] => 4 comments, 3 IRC mentions
[05:59:20]<hefring>memtkmcc: I'll pass that on when colan is around.
[07:03:46]<bgm>I'm playing on an old Aegir server (was on Aegir 1.x, then 2.x, now 3.x), and having trouble when cloning: d()->service('db')->connect(); returns " Access denied for user 'root'@'localhost' (using password: YES)". However, the credentials in ~/.drush/server_localhost.alias.drushrc.php are OK. Anywhere else I should check?
[07:16:52]<bgm>fixed by adding debug statements to dump the password used by aegir, then updated the mysql password and drush alias.
[07:18:22]<memtkmcc>bgm: Check db_passwd field in the hosting_db_server table, and make sure it matches what you have in server_localhost.alias.drushrc.php — both the password and the user. Note that the password is expected to be encoded, but should match. Also, it is possible that you have duplicate grants for root, like, different for localhost and different for 127.0.0.1, which can confuse mysql server.
[07:19:20]<memtkmcc>ah, ok
[07:21:07]<bgm>memtkmcc: ah, good to know, thanks!
[07:21:35]<bgm>(I hadn't checked there)
[07:35:34]<colan>memtkmcc: thanks for all of that analysis!
[07:35:34]<hefring>colan: 1 hour 36 min ago <memtkmcc> tell colan: https://www.drupal.org/node/2839973#comment-11846690
[07:36:40]<colan>i wonder where we should stick that code of yours.
[07:40:44]<memtkmcc>colan: it could be a separate function, not limited to the deploy code, but it is not going to save too many lines of duplicated code in contrib, since to be able to use it, you will have to add these 3 lines to include the correct file explicitly where needed, so I don't see too much reason to abstract it like in the proposed and non-working patch
[07:41:15]<colan>memtkmcc: so just stick the code directly in provision?
[07:41:31]<colan>that was my workaround anyway. :)
[07:42:22]<memtkmcc>yes, like in my example (used in BOA for a long time), since it also allows you to differentiate the if() filters, etc.
[07:48:00]<memtkmcc>colan: since contrib is outside of provision tree, it can't call functions in provision without including the file with the function, so it is easier to just code it directly where needed, especially if we couldn't really avoid any meaningful amount of code duplication by adding yet another hook, but of course if we want to prioritize code standards, we can do that, just not tied to deploy only, where it is
[07:49:56]<colan>makes sense.
[09:11:37]* memtkmcc has quit (Quit: Leaving.)
[10:12:37]* fatguylaughing has quit (Quit: fatguylaughing)