[14:15:59] Hi all, I created my own database and tried to use it in my python API but it is connecting to read only `_p` version of my database [17:11:34] Hi all, I created a new database s53794__leaderboard and created tables inside of it [17:11:55] but in python when I try to connect with it, I get access denied. [17:13:06] I am using the username and password from replica.my.cnf and hostname as enwiki.analytics.db.svc.eqiad.wmflabs [17:17:28] user created databases have hostname tools.db.svc.eqiad.wmflabs [17:19:58] Thanks a lot, this was the problem I was facing :) [18:25:18] Hello. [18:25:31] Please, create me a Toolforge account. [18:34:26] RhinosF1 Hello. [18:34:42] RhinosF1 Hello, i need Wikimedia Toolforge. [18:34:51] !log tools.zppixbot revoke access of EK for ToU vio [18:34:53] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zppixbot/SAL [18:34:58] !log tools.zppixbot-test revoke access of EK for ToU vio [18:34:59] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zppixbot-test/SAL [19:05:40] joaquinito01, hi [19:05:46] joaquinito01, did you try to log into toolsadmin.wikimedia.org ? [19:06:00] Krenair I cannot login, i don't have account. [19:06:02] I need account. [19:06:44] joaquinito01, did you click the sign up link and go through https://toolsadmin.wikimedia.org/register/ ? [19:07:03] Krenair I am banned. [19:07:33] Then we're probably not going to give you access. [19:07:43] What is your SUL account name? [19:07:53] Krenair: same name [19:08:32] Krenair Joaquinito01. [19:08:41] Yeah no definitely not going to approve an account if you have a globally locked account like that. [19:09:10] Krenair I am globally locked. [19:09:16] indeed. [19:09:28] why are you asking for toolforge access while being globally locked? [19:09:43] Krenair: its joaquinito01's thing [19:11:07] RhinosF1 I can use Toolforge??? [19:11:50] no [19:12:07] I'd talk to T&S first [19:12:27] RhinosF1 I have created a Oposta account. [19:14:54] RhinosF1 I have created a Oposta account. Request global lock. [19:15:08] Krenair: ^ [19:15:26] I'm not a steward [19:15:42] RhinosF1 I have created a Oposta account. [19:16:00] Urbanecm: ^ [19:17:03] RhinosF1 I have created a Oposta account. [19:17:43] thanks Krenair [19:18:05] I made a note in -stewards [19:18:22] Urbanecm: can you lock the wiki account? [19:19:50] done [19:31:21] Thanks [20:58:34] hi foks [21:00:03] er, actually, PM [21:02:08] foks: can you ack my ca@ email as well and give an idea on evidence needed pls? [21:02:48] RhinosF1, its the weekend :) [21:03:55] foks: I've lost many weekends and evenings to it as well. Quick peek? [21:04:04] Or from whoever is on call [21:05:16] we don't really have an on-call for tha [21:05:17] t [21:16:47] foks: ack [21:39:34] !log tools.zppixbot-test deploy a status config change (temp due to fallout) and add new twitter api key --T255922 [21:39:36] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zppixbot-test/SAL [21:39:39] !log tools.zppixbot deploy a status config change (temp due to fallout) and add new twitter api key --T255922 [21:39:40] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zppixbot/SAL [22:01:41] * RhinosF1 wonders why this is the time everything is slow [22:09:16] !help https://phabricator.wikimedia.org/T255926 [22:09:16] If you don't get a response in 15-30 minutes, please create a phabricator task -- https://phabricator.wikimedia.org/maniphest/task/edit/form/1/?projects=wmcs-kanban [22:12:38] RhinosF1, can't view it [22:13:22] Krenair: are you a cloud admin/root? [22:13:32] Krenair: I'm here, let me catch up a bit [22:14:26] RhinosF1, uh I'm a cloud-wide root but not admin [22:15:05] Krenair: not sure if you can do it, it should be accessible to acl*security though [22:15:26] (I can't add myself to random projects and start creating/deleting instances etc. for them but I can log into all instances as root, providing they're not totally broken) [22:15:44] oh and.rew is here, okay :) [23:05:57] !log tools.zppixbot-test restart all pods to unfallout T255926 [23:05:59] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zppixbot-test/SAL [23:08:32] !log tools.zppixbot restart all pods to unfallout T255926 [23:08:33] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zppixbot/SAL [23:12:40] I put "restartPolicy: Never" in the cronjob container spec, and when the container fails it don't restart, but the k8s cronjob create another pod with new container and start the same script again, how can I avoid it? [23:15:19] don't use a cronjob [23:15:52] what are you trying to do? [23:16:39] !log tools.zppixbot scrambled email&pw for a user [23:16:41] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.zppixbot/SAL [23:20:55] I am running a script that process a dump and update a database, when it fail I don't want it to restart to avoid introduce some error in the database [23:22:14] danilo: is it a one time thing? Or is it a job that you want to run on an automatic schedule? [23:22:39] I know I can run it in a single pod to debug, but I thought there was a simple way to avoid that restart [23:22:59] a "restart" is about a single job instance [23:23:07] bd808: it will run monthly [23:24:26] ok, I will run in a pod to debug before set the cronjob again [23:25:29] In Kubernetes, https://kubernetes.io/docs/concepts/workloads/controllers/job/ is the thing that can have a restartPolicy. And then https://kubernetes.io/docs/concepts/workloads/controllers/cron-jobs/ is a factory that creates Job objects on a schedule. [23:26:48] It sound like you are wanting a Job to tell it's owning CronJob to disable itself? [23:28:36] There is a flag for this in the CronJob object (.spec.suspend) but I'm not sure that there is a simple way to toggle the flag based on exit status of the last executed Job [23:35:57] bd808: I thing what I want is .spec.backoffLimit = 1, but read that in the "jobs" page, can I use it in cronjobs? [23:38:42] Yes, you can use that, but it applies to the Job that is spawned by the CronJob. It does not apply to the CronJob itself. [23:39:09] If you have a schedule of once a month, each month's run is a Job [23:39:51] CronJob creates Job based on spec in CronJob when schedule in CronJob matches current wall clock time. [23:41:12] if you are familiar with a unix cron system, CronJob is basically a crontab line. And Job is the process spawned by cron when the crontab line matches the system time. [23:43:35] I think I understand what you would like to happen, but I don't think Kubernetes has a built-in system to do that. Nor does a traditional unix cron (like vixiecron which is probably th emost commonly installed unix cron system). [23:43:40] yes, that is what I want, if the script fails in the monthly run I want it to not restart till the next month, so I can see and fix the error in that time [23:44:15] that should be the behavior if you set "restartPolicy: Never" [23:45:53] I set it in the container spec, the container didn't restart, but the job create more 3 pod that started and failed too [23:46:54] I saw some examples of the use of backoffLimit in cronjobs in stackoverflow, I will try that [23:47:45] hmmm... yeah. reading the Job documentation I think I thought "restartPolicy: Never" did something different than it actually does. [23:48:59] the easiest thing to do would probably be to use a lock file [23:49:49] it appears that "restartPolicy: Never" is about a single Pod created by the Job. But the Job still wants to get to the completions count and so will start new Pods rather than restarting the existing Pod. [23:50:21] this is going to confuse lots of people! [23:52:26] check if a file exists, if it doesn't, create it and run. when successful, delete the file. if it already exists, quit [23:56:36] I am getting "kubectl: command not found" [23:57:55] oh no, forget, I am in a shell inside a pod