| [18:05:05] | * Shwele has joined #aegir |
| [19:53:20] | * Shwele has quit (Ping timeout: 272 seconds) |
| [20:23:44] | * Shwele has joined #aegir |
| [22:00:02] | * shaneonabike has joined #aegir |
| [00:50:11] | * theMusician has joined #aegir |
| [01:12:13] | * Shwele has quit (Read error: Connection reset by peer) |
| [02:08:32] | * theMusician has quit (Ping timeout: 248 seconds) |
| [02:35:18] | <shaneonabike> | @colan[m]: are you around? |
| [02:42:33] | <colan[m]> | shaneonabike: yes. still need help? |
| [02:42:58] | <shaneonabike> | colan[m]: yes please |
| [02:43:15] | <shaneonabike> | colan[m]: if you don't mind |
| [02:43:28] | <colan[m]> | shaneonabike: sure, what's up? |
| [02:44:02] | <shaneonabike> | basically something went wrong on the latest aegir3 upgrade using the package manager |
| [02:44:36] | <shaneonabike> | In the main interface i can get to the platforms page |
| [02:44:53] | <colan[m]> | shaneonabike: error msgs? |
| [02:45:05] | <shaneonabike> | BUT when I get to any other page (like the site page or the listing of sites for a platform) I get he website encountered an unexpected error. Please try again later. |
| [02:45:23] | <shaneonabike> | colan[m]: my struggle is where exactly does Aegir output errors.. like it would normally be in error.log or it's own right |
| [02:46:31] | <shaneonabike> | OH! and colan[m] the latest hostmaster is live but not in teh web interface |
| [02:46:59] | <shaneonabike> | colan[m]: my aegir hostmaster moved just fine but the thing is muddled |
| [02:48:12] | <colan[m]> | shaneonabike: right. typically these would be in the drupal log, so enable the Database Logging module, and then surf to ... |
| [02:48:46] | <colan[m]> | Reports -> Recent log messages. |
| [02:49:01] | <shaneonabike> | colan[m]: right i enabled syslog give me a sec to check |
| [02:49:09] | <shaneonabike> | colan[m]: i guess i assumed it was just enabled whoops |
| [02:49:28] | <colan[m]> | syslog works too, but they're both off by default. ;) |
| [02:49:56] | <colan[m]> | alternately, from the CLI, you can do `drush ws --tail --full` and keep it running there while you do stuff. |
| [02:50:24] | <colan[m]> | er, `drush @hostmaster watchdog-show --tail --full` |
| [02:50:42] | <shaneonabike> | ws worked ;) |
| [02:51:04] | <shaneonabike> | Error 'Class 'ResultSet' not found in hosting_report_node_load |
| [02:51:56] | <colan[m]> | cool. so basically, if drupal loads pages, and it tells you you have an error, check there. if it doesn't load, WSOD for example, then check web server logs (nginx or apache). |
| [02:53:14] | <shaneonabike> | When I load the error page i get Error: Class 'ResultSet' not found in hosting_report_node_load() (line 131 of /var/aegir/hostmaster-7.x-3.175/sites/domain/modules/contrib/hosting_report/hosting_report.module |
| [02:53:32] | <shaneonabike> | That's pretty much the same issue everywhere |
| [02:54:24] | <colan[m]> | shaneonabike: ok. so if that's not a local problem (e.g. php files were deleted, module is missing, etc.), open an issue in that queue: hosting_report_node_load |
| [02:54:54] | <colan[m]> | there's nothing related to that there now, just checked. |
| [02:55:55] | <colan[m]> | (might also be a permissions issue) |
| [02:56:12] | <shaneonabike> | This has been working for over a year so that seems weird |
| [02:56:25] | <shaneonabike> | Unless something got ubber messed.. I had checked the permissions in the systme and they appear "ok" |
| [02:56:46] | <colan[m]> | shaneonabike: sure, but nobody else has run into it yet, it seems. |
| [02:57:00] | <shaneonabike> | egads not fun |
| [02:57:09] | <shaneonabike> | is there a way i could just rollout another hostmaster ? |
| [02:57:23] | <shaneonabike> | platform that is and then migrate my site |
| [02:58:00] | <colan[m]> | shaneonabike: well, besides a normal installation, you can use https://gitlab.com/aegir/aegir-dev-vm to test things. |
| [02:59:45] | <colan[m]> | a known related error is https://www.drupal.org/project/hosting_site_audit/issues/2882493, for which i disable the module before upgrades. but this looks different. |
| [02:59:46] | <hefring> | https://www.drupal.org/project/hosting_site_audit/issues/2882493 => Debian package upgrades fail with "redeclaration of probes_site_audit_drush_command()" [#2882493] => 4 comments, 1 IRC mention |
| [03:01:06] | <colan[m]> | not sure you can have two hostmaster on the same install. probably not a good idea. how about comparing the module code from the old one? a diff should tell you what's missing. |
| [03:01:21] | <shaneonabike> | no i only have one |
| [03:01:36] | <shaneonabike> | and that one migrated properly to the correct hostmaster |
| [03:01:51] | <shaneonabike> | but what I wanted to point out is that the new hostmaster isn't in the aegir interface UI |
| [03:02:01] | <shaneonabike> | is there a way I could force an import to see if that might the issue?? |
| [03:02:56] | <colan[m]> | i would probably try `drush @hm provision-verify` |
| [03:04:28] | <colan[m]> | i've run into that before, but not sure i recovered. think i just set up a new install, and Remote(ly) Import(ed) the sites to the new one. this was an install i took over so i didn't want to trust the old one anyway. |
| [03:04:57] | <shaneonabike> | Hmm in this case it says that @hm and @hostmaster don't exist |
| [03:05:02] | <shaneonabike> | So I think we are getting somewhere |
| [03:05:14] | <shaneonabike> | they are in teh drush aliases tho |
| [03:09:17] | <colan[m]> | good. not sure i can do much more without looking at your system. if you get stuck, maybe we can jump in with some paid help if you need to dig deeper. |
| [03:09:56] | <colan[m]> | ACTION goes to grab some lunch. |
| [03:29:38] | <shaneonabike> | i disabled hosting monitor and the problem is gone hrmm |
| [03:48:11] | <colan[m]> | shaneonabike maybe try a reinstall? |
| [03:48:32] | <shaneonabike> | actually i created the platform and it recognized it, modified the db, and now it's working |
| [03:48:40] | <shaneonabike> | something is weird with the two modules |
| [03:48:58] | <shaneonabike> | now I noticed that hosting-queued is not working |
| [03:49:15] | <shaneonabike> | this makes me a bit concerned: |
| [03:49:15] | <shaneonabike> | Jun 17 13:45:24 aegir.can-latam.org systemd[1]: hosting-queued.service: Failed with result 'exit-code'. |
| [03:49:15] | <shaneonabike> | Jun 17 13:45:31 aegir.can-latam.org sshd[4424]: Connection closed by 218.92.0.194 port 20317 [preauth] |
| [03:49:15] | <shaneonabike> | Jun 17 13:45:31 aegir.can-latam.org sshd[4425]: Invalid user qiao from 51.38.90.195 port 57990 |
| [03:49:15] | <shaneonabike> | Jun 17 13:45:31 aegir.can-latam.org sshd[4425]: input_userauth_request: invalid user qiao [preauth] |
| [03:49:16] | <shaneonabike> | Jun 17 13:45:31 aegir.can-latam.org sshd[4425]: pam_unix(sshd:auth): check pass; user unknown |
| [03:57:07] | <shaneonabike> | colan[m]: actually i see on other servers this doesn't work. Can you just help me identify where the cron job is being launched from |
| [03:57:13] | <shaneonabike> | colan[m]: for aegir at least |
| [03:58:23] | * theMusician has joined #aegir |
| [03:59:26] | <colan[m]> | I think it's in the Aegir user's crontab. But it's best restarted with `sudo service hosting-queued restart` |
| [04:01:23] | <colan[m]> | shaneonabike it's not a great daemon as it can die frequently. Skynet is an alternative. |
| [04:02:10] | <shaneonabike> | colan[m]: yeah it's telling me that it caned find drush hosting-queued |
| [04:02:19] | <shaneonabike> | when i start it which is odd |
| [04:02:52] | <shaneonabike> | actually when I test what it is trying to do to start |
| [04:03:33] | <colan[m]> | Hmm |
| [04:03:58] | <shaneonabike> | colan[m]: like if i do /usr/local/bin/drush --debug @hostmaster hosting-queued |
| [04:04:04] | <colan[m]> | Look in the crontab |
| [04:04:22] | <shaneonabike> | yeah i checked that before ther was nothing |
| [04:04:46] | <shaneonabike> | viola |
| [04:05:03] | <shaneonabike> | i see on my other server it had a crontab and for whatever reason this has been removed |
| [04:05:04] | <shaneonabike> | good lord |
| [04:06:05] | <colan[m]> | Yeah, that should be there. |
| [04:06:17] | <shaneonabike> | i copied it over |
| [04:11:31] | * shaneonabike has left #aegir () |
| [04:38:51] | * shaneonabike has joined #aegir |
| [04:39:43] | <shaneonabike> | colan[m]: now when I do a provision-verify @hm it says Checking DB credentials yielded error: Unable to find a matching SQL [debug] |
| [04:39:43] | <shaneonabike> | Class. Drush cannot find your database connection details. |
| [04:40:20] | <shaneonabike> | colan[m]: i can however do a drush sqlc and connect. SO before I pull my hair out I was wondering if it's safe to just move the hostmaster platform (leave all the other sites) and basically just run the hosmaster-install again |
| [04:40:34] | <shaneonabike> | colan[m]: i would imagine it should just "identify" all the platforms that exist right? |
| [04:41:04] | <shaneonabike> | I've used aegir for like 6 years I feel and never had this kind of weirdoness |
| [04:43:12] | <colan[m]> | I'm afraid I can't tell you if that's safe or not, shaneonabike , as I've never tried that. |
| [04:43:44] | <colan[m]> | I wouldn't rely on "should" here. :) |
| [04:43:52] | <shaneonabike> | colan[m]: haha |
| [04:44:05] | <shaneonabike> | colan[m]: yeah... but I don't think i can rebuild the hostmaster can i |
| [04:45:54] | <colan[m]> | not that i'm aware. snapshot the VM before trying? i'd probably just import the sites into a new install (if all else fails). |
| [04:47:48] | <shaneonabike> | silly question here but would the php version matter |
| [04:47:50] | <shaneonabike> | i'm on php7.3 |
| [04:48:03] | <shaneonabike> | 7.3 fpm |
| [04:55:19] | <shaneonabike> | colan[m]: thanks anyway for all your help |
| [04:55:21] | * shaneonabike has left #aegir () |
| [08:23:51] | <colan[m]> | don't think i'm running that anywhere so no idea, sorry! |
| [08:50:06] | * shaneonabike has joined #aegir |
| [08:57:29] | * shaneonabike has left #aegir () |
