[09:29:24] <dcaro>	 !status paws down
[09:30:47] <dcaro>	 !paws reboot worker 1
[09:31:07] <dcaro>	 hahah
[09:31:10] <dcaro>	 !log paws reboot worker 1
[09:31:11] <dcaro>	 xd
[09:31:14] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Paws/SAL
[09:32:37] <dcaro>	 !log paws reboot worker 2 and 4
[09:32:38] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Paws/SAL
[09:34:41] <dcaro>	 !status OK
[13:17:32] <Guest09494>	 Hello there! @dhinus. I have made a membership request on Toolforge on 12 th of November and got a response on 13th November, it was normal but then when I respond back because they need some kind or information, so from that day it had been almost 6 to 7 days I got no response. It is a humble request to check it once again. Thanks!
[13:18:41] <dhinus>	 dcaro: can you have a look? ^
[13:20:23] <Guest09494>	 Hello there! @dhinus. I have made a membership request on Toolforge on 12 th of November and got a response on 13th November, it was normal but then when I respond back because they need some kind or information, so from that day it had been almost 6 to 7 days I got no response. It is a humble request to check it once again. Thanks!
[13:21:00] <taavi>	 Guest09494: you do not need to send the exact same message here multiple times.
[13:22:12] <Guest09494>	 Thanks I don't I hope my request is reached safely :)
[13:23:21] <dcaro>	 I'm not in front of a laptop, I'll be in 20min
[16:03:31] <andrewbogott>	 !log tools increased object storage quota to 200GB
[16:03:34] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL
[16:08:49] <andrewbogott>	 isaacj: ok if I reboot content-diff-index.wmf-research-tools ?
[16:11:41] <andrewbogott>	 mgerlach, same question ^
[16:19:16] <andrewbogott>	 !log tools increased object count quota to 100,000
[16:19:18] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL
[16:23:56] <andrewbogott>	 xtex, are you 'XtexChooser'? If so, would it be OK for me to reboot the 'journalist1' VM in the wmgmc-monitoring  project?
[18:55:09] <wm-bot>	 !log lucaswerkmeister@tools-bastion-15 tools.lexeme-forms deployed 7f658a6675 (update Bootstrap)
[18:55:12] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.lexeme-forms/SAL
[19:01:36] <wm-bot>	 !log lucaswerkmeister@tools-bastion-15 tools.speedpatrolling deployed e6d93e0064 (Bootstrap 5.3)
[19:01:38] <stashbot>	 Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.speedpatrolling/SAL
[19:15:01] <isaacj>	 andrewbogott: all good to reboot. thanks!
[19:42:52] <andrewbogott>	 great, thanks for checking
[19:53:25] <derenrich>	 if i run `toolforge webservice shell` is that pod supposed to get automatically cleaned up? or is that on me
[19:55:34] <wm-bb>	 <lucaswerkmeister> it should get cleaned up as soon as you exit the shell
[20:02:23] <derenrich>	 hmm that's not happening for me
[20:02:33] <derenrich>	 maybe because my connection breaks?
[20:04:07] <wm-bb>	 <lucaswerkmeister> hm, not sure what happens in that case
[20:04:12] <wm-bb>	 <lucaswerkmeister> which tool is this in?
[20:04:27] <derenrich>	 best-of. but i just deleted a bunch of them
[20:06:24] <wm-bb>	 <lucaswerkmeister> ok, yeah, looks like the pod might stay alive if the connection is terminated
[20:06:35] <wm-bb>	 <lucaswerkmeister> (tested with `~.` in SSH)
[20:07:32] <wm-bb>	 <lucaswerkmeister> I wonder if we should set the `$TMOUT` variable inside the shell so bash will exit after a while
[20:07:45] <wm-bb>	 <lucaswerkmeister> (though I’ve been bitten by that behavior before, so I’m not *convinced* it’s a good idea)
[20:45:01] <derenrich>	 i would be fine if those pods just got reaped 
[20:45:09] <derenrich>	 but idk what the average person uses them for
[20:47:56] <wm-bb>	 <lucaswerkmeister> well, hopefully nothing non-interactive, that’s not what they’re meant for ;)
[20:51:04] <derenrich>	 right
[20:52:25] <bd808>	 @lucaswerkmeister: there is no control of the image used in a `webservice shell` run so $TMOUT would be challenging to implement even if it wasn't the most evil thing in shells.
[20:55:48] * lucaswerkmeister confused
[20:56:31] <lucaswerkmeister>	 we control which images `webservice` maps the types to, and the build process for those images, and the container spec for the pod (e.g. we could mount stuff over /etc/profile)
[20:56:38] <lucaswerkmeister>	 surely that’s a lot of control?
[20:56:49] <lucaswerkmeister>	 not that users couldn’t circumvent it, of course, but that’s not the intention ^^
[20:57:02] <lucaswerkmeister>	 but sure, I’m also okay with saying $TMOUT is to be avoided
[21:02:55] <taavi>	 tbh we could just label each shell pod with the bastion it was launched from, and then run a job killing all shell pods without a matching `kubectl` process on that bastion
[21:21:09] <bd808>	 @lucaswerkmeister: the image could just as easily be a build service created container under the control of the tool maintainer or another arbitrary tool maintainer. We are currently using unmodified base images from the upstream project for these. This is why I said we have no control.
[21:21:41] <lucaswerkmeister>	 ah… I was just thinking of non-buildservice types
[21:22:02] <lucaswerkmeister>	 not sure I even realized you can use webservice shell with buildservice tools… good to know 🤔
[21:24:23] <bd808>	 @lucaswerkmeister:  you can even make your own custom bastion with one -- https://gitlab.wikimedia.org/toolforge-repos/bd808-buildpack-perl-bastion
[21:25:33] <bd808>	 That repo also answers the question "how can I run arbitrary code during the buildpack build process"