[10:11:32] hi, i am having problems with my nullpool implementation with flask sql alchemy. Every now and then the connection to the database braks. Is there an example implementation or a toolforge project that has a nullpool implementation that i could look at? [10:13:24] Guest21: I’m not sure what a nullpool is, but I’ll just mention that persistent connections are not permitted, since it sounds like you might be trying to do something like that https://wikitech.wikimedia.org/wiki/Help:Toolforge/Database#Connection_handling_policy [10:14:04] (hm, though https://docs.sqlalchemy.org/en/14/core/pooling.html suggests nullpool is how you *disable* pooling, so maybe you’re trying to do the right thing?) [13:54:36] !log tools enable distro-wikimedia unattended upgrades T290494 [13:54:38] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [13:54:38] T290494: Revisit Toolforge automated package updates and version pinnings - https://phabricator.wikimedia.org/T290494 [13:56:48] !log paws upgrade pywikibot on container start 437f46a1caf56cf860f85211af4567735410a621 [13:56:49] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Paws/SAL [13:59:11] !log admin manually attached [[User:Dreamy Jazz]] to wikitech for a password reset (https://wikitech.wikimedia.org/wiki/Portal:Toolforge/Admin#Manually_associate_an_LDAP_account_with_wikitech) [13:59:12] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [15:18:29] !log tools depool tools-k8s-worker-42 for experiments [15:18:31] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [15:26:04] ^ uncordoned [15:41:16] !log rebooting cloud*-dev for T307668 [15:41:17] andrewbogott: Unknown project "rebooting" [15:41:26] !log admin rebooting cloud*-dev for T307668 [15:41:27] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [15:42:31] ^ does that include cloudbackup1*-dev stuff? [15:42:53] (and not cloudbackup2*?) [15:55:25] dcaro: yes [16:06:00] okok, just making sure (I have made that mistake at some point) [16:13:39] dcaro: I'm trying to get the rabbit cluster to stand back up on cloudcontrol200x-dev. I'm unclear what's happening so far, 2004 says it can't reach 2001 [16:14:12] andrewbogott: looking [16:14:23] docs for this are approximately https://wikitech.wikimedia.org/wiki/Portal:Cloud_VPS/Admin/Rabbitmq [16:15:33] doesn't seem to be firewall although I always suspect ipv4 vs ipv6 firewall things [16:19:38] I see this in the logs: 2022-05-10 16:17:58.306 [warning] <0.274.0> Error while waiting for Mnesia tables: {timeout_waiting_for_tables,['rabbit@cloudcontrol2003-dev'],[rabbit_queue,rabbit_durable_queue]} [16:19:54] seems like it fails before starting up, trying to load the DB from disk, maybe it got corrupted? [16:21:55] hmm... no, that seems related to the clustering [16:22:15] yeah, I already went through a cycle of trying to delete the mnesia tables and start over. Possibly I made it worse, but that's the same issue I saw before I stopped everything and wiped out the tables [16:23:12] It's probably worth doing apt --reinstall and letting it regenerate all of the support files and such in case there's still corruption someplace [16:23:19] but I won't do that until you're standing clear :) [16:23:28] okok, I'm fooling around a bit [16:24:25] interesting: AMQPLAIN login refused: user 'nova' - invalid credentials [16:26:32] oh... maybe my attempt to clean up wiped out the nova account? But that shouldn't prevent clustering [16:37:53] dcaro: ready for me to wipe things out or are you still digging? [16:37:56] * epmd reports node 'rabbit' uses port 25672 for inter-node and CLI tool traffic [16:37:58] * can't establish TCP connection to the target node, reason: timeout (timed out) [16:38:30] I mean, I'm ready xd [16:38:43] hmmmm maybe it is firewall? [16:39:19] 25672 is not in the firewall [16:39:21] that port isn't mentioned in ferm at all [16:41:15] 5672 is there though [16:41:38] yep, it was a mistake in a diff reshuffling things [16:41:44] it was there too [16:43:08] andrewbogott: for your review https://gerrit.wikimedia.org/r/c/operations/puppet/+/790725 [16:43:23] (/me looking if there's any other missing ports from there xd) [16:43:46] https://gerrit.wikimedia.org/r/c/operations/puppet/+/787003 xd [16:43:50] it had a comment ;) [16:44:00] but in draft... xd [16:44:47] oh dang [16:45:00] well, that explains a lot :) [16:45:05] I think that until you reboot ferm does not remove rules [16:45:17] (similar thing happened with cloudnets) [16:45:24] yeah, that totally fits everything [16:45:36] go ahead and merge? And then I'll try to repair whatever damage I did while troubleshooting [16:45:52] 👍 [16:46:21] you want me to run puppet on the hosts? or should I wait until you get rabbit sorted out? [16:48:41] I'm doing it [16:48:47] ack [16:49:26] oh, you rebooted something? [16:49:49] (lost connection to cloudcontrol2004-dev) [16:50:13] oh yeah, sorry -- just trying to get things back into an expected state (it was complaining about mnesia not running) [16:51:25] no prob [16:51:42] fyi. 2003 still does not have the fw rules (2001 does) [16:52:00] oops. sorry about that! [16:52:20] what happened? [16:52:43] missing that port when shuffling things [16:53:01] (just now reading the backlog and noticed that it was my patch that broke things) [16:54:09] aaahhh, no problem :), thanks a lot for the patch, stuff happens, keep them coming ;) [16:54:38] 2004 is working now, 2003 is having some sort of fit, I might just reboot it too if that won't mess with you [16:54:54] andrewbogott: it does not have the firewall rules applied, did puppet run? [16:55:03] puppet won't run [16:55:06] I mean, it h angs [16:55:11] oh [16:55:13] that's not nice [16:55:23] I don't see a puppet process [16:55:34] and now I'm terminal there is getting all jittery [16:55:45] ok if I reboot? [16:55:46] works ok for me [16:55:51] ok for reboot too [16:56:07] * andrewbogott type 'reboot' sees 'ret' in the terminal [16:56:25] hmmm... maybe you have connectivity issues? [16:56:33] (or the path your ssh connection takes) [16:58:17] I was able to run puppet, and it applied the ferm rule [16:58:32] (took a long time trying to start rabbit first though, maybe some ordering would help) [16:58:35] dunno, all my other terminals (on other cloudcontrols) were still responsive [16:59:01] hmm... [16:59:01] oh, heh, I thought it rebooted but my terminal messed up the mollyguard thing [16:59:13] it's behaving ok for the moment [16:59:44] Yep, I see it. [16:59:49] okok [16:59:54] This is where I'm getting stuck: [16:59:56] the machine seems ok to me [16:59:57] https://www.irccloud.com/pastebin/QozJdT5o/ [17:00:03] but I'll just be patient [17:01:03] aaaahhh, yep, that will probably not work, you can try stopping the service and starting it again though, itś strying to start, on the `waiting for Mnesia` bit [17:01:12] 4 more tries until it gives up [17:01:21] (2min) [17:01:31] ok [17:01:51] and now my terminal is messed up again [17:01:59] that is weird :/ [17:02:12] 2001 and 2004 see each other though, that's good [17:02:48] ok, there, I finally got it to reboot [17:02:53] perhaps it will make sense after that [17:03:32] :crossingfingers: [17:07:37] it did not join the cluster yet [17:08:00] same issue [17:08:05] (it seems) [17:10:05] yeah, I'm trying to reinstall the package but it wont' reinstall until it can stop and it's trying... [17:12:29] okok, probably zapping it (reset) and forcing it to re-join might be enough [17:14:35] I can't get it to do anything but hang but I'll let you give it a try :) [17:15:22] okok, on it [17:20:08] oooohhh, I got the same issue with the console, when I Ctr+C while stopping the app [17:20:24] that fits with what I saw [17:20:30] interesting [17:20:56] reset did not help [17:22:15] maybe kill everything, rm -f all files, reinstall [17:22:25] (although I don't know what 'all files' consists of specifically) [17:25:04] ok, I moved the mnesia tables out, the server started, now we can reset it and rejoin [17:25:15] great, are you doing or shall I? [17:25:21] on it :) [17:25:24] cool [17:26:50] okok, so the three are up [17:27:02] still getting the credentials error xd [17:27:38] I think puppet might be the one creating the users/auth? [17:28:27] Notice: /Stage[main]/Openstack::Neutron::Rabbit/Rabbitmq::User[neutron-rabbituser]/Exec[rabbit_neutron_setup_perms]: Triggered 'refresh' from 1 event [17:28:39] it's creating stuff [17:29:48] okok, the errors stopped [17:30:05] We might still need to create a trove user by hand for VM workers [17:30:08] but otherwise I think we're good [17:30:37] and I see no errors on the journal logs (from openstack stuff) either [17:30:41] looks ok yep :) [17:31:33] okok, gtg, let me know at the end of the day if you want me to continue anything tomorrow [17:31:35] yep, creating [17:31:45] ok! I need lunch and then will do some more reboots [17:31:53] thank you for getting 2003 unstuck! [17:31:56] good luck! [17:32:03] np! [17:48:58] Question re: -- is it possible to proxy multiple backend ports? [17:49:18] or is it one port per hostname? [17:53:53] ori: one proxy hostname (something.wmcloud.org/wmflabs.org) is linked to one backend (instance+port), but you can have multiple proxy hostnames pointing to the same instance on different ports [17:55:11] ok, that's what I figured. thanks [17:58:12] Possibly a stupid question - but I've setup a cloud vps instance and I'm not sure which folder the default nginx install is using to serve web pages [17:58:50] It was /var/www/html/ in debian-10.0-buster [17:58:57] but not sure where it is on debian-11.0-bullseye [17:59:25] Jdlrobson: maybe check the /etc/nginx/... config files? [17:59:50] I can't locate the nginx folder either [18:00:17] hmmm... but you did apply a puppet role or manually install nginx? [18:00:48] puppet role. [18:01:54] i haven't setup nginx manually but traffic seems to be going to nginx on https://pixel2.wmcloud.org/ [18:02:16] that's probably the front-proxy nginx [18:02:50] ah okay, so it's not installed. I'm guessing in debian-10.0-buster it was? [18:02:58] That would explain my confusion :) [18:03:24] (in the past I've simply set up vps instances and had a static server without any additional work) [18:04:18] Not sure how that would have happened for you honestly Jdlrobson unless there is some prefix puppet you have setup in the reading-web-staging project. [18:04:34] there is no default httpd in the base images we make available [18:04:39] Okay thanks for clearing that up [18:04:59] I don't see any puppet config at all on https://openstack-browser.toolforge.org/server/pixel.reading-web-staging.eqiad1.wikimedia.cloud which is the backend that pixel2.wmcloud.org points at. [18:05:02] When I setup skins.wmflabs.org/ it just worked out the box which may be very happy [18:05:10] I'm sure not sure that worked though [18:05:53] we're using es-dev-server for pixel but it doesn't survive reboots so I was curious what magic was powering skins.wmflabs.org [18:05:56] Jdlrobson: you have role::simplelamp2 applied on the backend host for skins -- https://openstack-browser.toolforge.org/server/skins.reading-web-staging.eqiad1.wikimedia.cloud [18:06:18] ahhh [18:06:51] oh, someone uses it and was happy about it, yay [18:07:03] mutante: you made role::simplelamp2 ? [18:07:09] or added ? [18:07:13] yea, it was to replace simplelamp [18:07:35] well it's working so well i didnt even know i was using it :) [18:07:43] okay that was the missing piece of the puzzle i needed! [18:07:44] :) [18:07:47] thank you bd808 [18:07:51] and mutante for putting it there :) [18:08:07] yw [18:08:31] you can use prefix/project puppet if you want it on every instance in a certain project [18:09:07] When making puppet changes I have to soft reboot? [18:10:23] no, just run the puppet agent or wait [18:10:48] sudo puppet agent -tv [18:11:35] `sudo run-puppet-agent`* [18:11:37] Might be worth adding a page for static hosting on https://wikitech.wikimedia.org/w/index.php?search=static+hosting&title=Special%3ASearch&go=Go&ns0=1&ns12=1&ns116=1&ns498=1 ? [18:15:36] technically it's not just static because this also installs PHP but I can't think of a better name [18:16:05] We are in general light on Cloud VPS specific Help: documentation. I agree that a little page on setting up a LAMP stack on an instance would be helpful. [18:16:32] oh... https://wikitech.wikimedia.org/wiki/Help:LAMP_instances [18:16:38] maybe needs some SEO help :) [18:17:28] found that with https://wikitech.wikimedia.org/w/index.php?search=httpd&title=Special:Search&profile=advanced&fulltext=1&ns12=1 [18:18:40] simplelamp2 (httpd, memcached, PHP, mariadb), simplelap (same but without mysql/maria) and simplestatic (really just httpd) are 3 different roles available [18:18:48] that page is nice [18:19:04] we currently have a technical writer working on docs by the way [18:19:23] they were converting a "how to create VMs in cloud VPS" google doc [18:36:28] sweet [18:52:00] does the proxy always map port 80 to the backend port? can i proxy a different port? [18:52:16] right now i have public-hostname:80 proxying private-instance:6927 [18:52:25] i'd like to have public-hostname:6927 proxy private-instance:6927 [18:53:35] not possible at the moment, only https on port 443 is proxied [18:53:51] err, :443, right [18:53:56] ok, I can live with that, thanks [18:54:30] one more questionm re: https://wikitech.wikimedia.org/wiki/Help:Using_a_web_proxy_to_reach_Cloud_VPS_servers_from_the_internet#Migrate_from_a_*.wmflabs.org_proxy_to_a_*.wmcloud.org_proxy [18:54:43] are *.wmflabs.org proxy addresses deprecated / slated for eventual deletion? [18:54:56] is there any advantage to using *.wmfcloud.org? [18:55:01] *wmcloud rather [18:55:11] yes, but we don't have any concrete plans yet for removing the old ones [18:56:16] "labs" is a horrible term that we've been trying to get rid of since mid 2017 https://wikitech.wikimedia.org/wiki/Help:Labs_labs_labs [18:57:08] yeah i know :) [18:57:10] thanks [18:57:24] * andrewbogott waves to ori [18:57:39] hello! [19:05:38] !log admin Safe rebooting 'cloudvirt1017.eqiad.wmnet'. - cookbook ran by andrew@buster [19:05:38] !log admin Draining 'cloudvirt1017.eqiad.wmnet'. - cookbook ran by andrew@buster [19:05:40] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:05:42] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:06:25] !log admin Set cloudvirt 'cloudvirt1017.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [19:06:26] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:29:44] !log admin Drained 'cloudvirt1017.eqiad.wmnet'. - cookbook ran by andrew@buster [19:29:47] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:33:06] !log admin Unset cloudvirt 'cloudvirt1017.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [19:33:06] !log admin Safe reboot of 'cloudvirt1017.eqiad.wmnet' finished successfully. - cookbook ran by andrew@buster [19:33:08] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:33:09] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:33:22] didn't we move all this spam to -feed? [19:36:30] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:36:30] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:36:33] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:36:34] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:37:15] !log admin Set cloudvirt 'cloudvirt1021.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [19:37:17] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:39:26] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:39:26] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:39:27] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:39:28] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:40:41] !log admin Set cloudvirt 'cloudvirt1021.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [19:40:43] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:41:02] taavi: the patch for that was merged, yes. Not sure that there were any tests to prove it worked though... [19:44:02] !log admin Safe rebooting 'cloudvirt1022.eqiad.wmnet'. - cookbook ran by andrew@buster [19:44:02] !log admin Draining 'cloudvirt1022.eqiad.wmnet'. - cookbook ran by andrew@buster [19:44:04] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:44:05] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:45:02] ori: $NAME.wmcloud.org proxies will also resolve as $NAME.wmflabs.org for backwards compat when migrating. We mean to disable creating new *.wmflabs.org names soon. (We actually just forgot that we had not previously.) [19:45:54] !log admin Safe rebooting 'cloudvirt1022.eqiad.wmnet'. - cookbook ran by andrew@buster [19:45:54] !log admin Draining 'cloudvirt1022.eqiad.wmnet'. - cookbook ran by andrew@buster [19:45:55] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:45:57] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:46:34] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:46:34] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:46:35] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:46:36] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:47:24] !log admin Set cloudvirt 'cloudvirt1021.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [19:47:26] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:55:54] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:55:54] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:55:56] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:55:57] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:57:27] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:57:27] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [19:57:29] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [19:57:30] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [20:00:31] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [20:00:31] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [20:00:32] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [20:00:37] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [20:01:57] !log admin Safe rebooting 'cloudvirt1023.eqiad.wmnet'. - cookbook ran by andrew@buster [20:01:57] !log admin Draining 'cloudvirt1023.eqiad.wmnet'. - cookbook ran by andrew@buster [20:01:59] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [20:02:01] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [20:05:09] !log admin Set cloudvirt 'cloudvirt1023.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [20:05:10] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [20:21:50] taavi: did you just create new instances? in that admin-monitoring project? [20:22:11] ah, the user name is right there. nevermind me [20:57:15] !log admin-monitoring cleanup fullstack test instances after cloudcontrol reboots [20:57:18] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin-monitoring/SAL [21:32:32] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [21:32:32] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [21:32:35] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [21:32:36] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [21:35:28] !log admin Set cloudvirt 'cloudvirt1021.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [21:35:29] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [21:40:05] !log admin Safe rebooting 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [21:40:05] !log admin Draining 'cloudvirt1021.eqiad.wmnet'. - cookbook ran by andrew@buster [21:40:07] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [21:40:08] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [21:43:02] !log admin Set cloudvirt 'cloudvirt1021.eqiad.wmnet' maintenance. - cookbook ran by andrew@buster [21:43:06] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL