[05:57:13] <hashar>	 claime: I have another oneliner for your breakfast: `tar -xzf artifacts/artifacts.buster.tar.gz  --to-command='printf "%s %s\n" "$(md5sum -z)" "$TAR_FILENAME"'`
[06:44:40] <_joe_>	 hashar: usually such things should be see *after* digestion is complete
[07:14:14] <hashar>	 happy to see you liked it =)
[07:40:20] <claime>	 Gave a good kick to my morning coffee thanks
[07:44:34] <claime>	 hashar: You'll be happy to know that doesn't work on system tar on macOS :)
[07:47:27] <claime>	 hashar: diff -u <(gtar -xzf <(git show HEAD^:artifacts/artifacts.buster.tar.gz) --to-command='printf "%s %s\n" "$(md5sum -z)" "$TAR_FILENAME"')  <(gtar -xzf /Users/claime/wmf/gerrit/docker-pkg-deploy/artifacts/artifacts.buster.tar.gz --to-command='printf "%s %s\n" "$(md5sum -z)" "$TAR_FILENAME"')
[07:47:35] <claime>	 Fun for the whole family right there
[07:47:51] <claime>	 (but gnu-tar specific)
[07:48:02] <hashar>	 doesn't surprise me :)
[07:49:26] <claime>	 And at the above oneliner point, I'm this close to just making a manifest, because sorting inside a makefile inside a subshell inside a tar --to-command starts to get iffy
[07:52:19] <hashar>	 I thought about upstreaming those commands to the python-build container
[07:52:31] <hashar>	 so they can be shared between downstream repos
[07:53:09] <claime>	 Makes sense, rather than putting them in the Makefile of docker-pkg-deploy like I did :p
[08:25:22] <hashar>	 for the scope of deploying docker-pkg 3.0.3 it is fine
[08:25:38] * hashar grabs a coffee before the reviews marathon
[09:12:04] <question_mark>	 we don't seem to have a presentation for the upcoming Monday meeting yet
[09:12:31] <question_mark>	 is anyone willing and ready to give a presentation perhaps? or maybe one of the few things that got punted from the SRE Summit's schedule last week?
[09:51:44] <Emperor>	 I'm part-way through re-writing the dgit one I was going to give...
[09:52:57] <question_mark>	 Emperor: i'm guessting that means "preferably not yet"
[09:56:13] <Emperor>	 I'd rather go for meeting N+1, but I could try and finish up today if you're really keen
[10:51:56] <question_mark>	 no I think that's ok
[10:52:08] <question_mark>	 please add your presentation to the backlog in the meeting notes though :)
[10:55:00] <Emperor>	 done
[13:20:31] <topranks>	 question_mark: I didn't get a chance to do the eqiad expansion review at the summit so I added it to the list too
[13:20:55] <topranks>	 I ceased all prep work once it had to be moved however, so Monday is probably a little soon - can do following week 10th Oct if that works
[13:39:13] <topranks>	 sry that should be 17th - and nabbed by Matthew.  and I'm out on 31st, so Nov. it'll have to be.
[13:47:04] <Emperor>	 topranks: if you want to do it sooner, I'd be OK to swap
[13:47:35] <topranks>	 Emporer: no particular rush on my side, but happy to swap if that suits you better 
[13:47:55] <Emperor>	 Nah, I'll only keep procrastinating finishing this talk off /o\
[13:48:31] <topranks>	 haha I'm in the same boat... but on the list now which is something :P
[14:31:28] <bblack>	 Friday interesting reading: https://www.pathsensitive.com/2022/09/bet-you-cant-solve-these-9-dependency.html
[14:31:40] <bblack>	 [all about what dependency really means in software]
[16:26:55] <diskdance[m]>	 can somebody please take a look at https://gerrit.wikimedia.org/r/c/operations/puppet/+/816161 ? thanks
[16:27:09] <diskdance[m]>	 which has been there for like 2 months
[16:50:58] <legoktm>	 diskdance[m]: sure, I've added it to my queue
[17:42:38] <bblack>	 TIL: https://en.wikipedia.org/wiki/Wikipedia:Reliable_sources/Perennial_sources#Wikidata (apparently wikidata transclusions have some questionability as fact sources, according to the wikipedia editor community)
[17:43:01] <bblack>	 "There has been controversy over the use of Wikidata in the English Wikipedia due to its infancy, its vandalism issues and its sourcing"
[17:43:32] <bblack>	 I ended up there because of the link to that page from https://slate.com/technology/2022/09/wikipedia-fox-news-reliability.html which in turn I found on HN today
[17:44:19] <bblack>	 that reliable sources page is fascinating all around.  it's a distillation of so many heated community debates about all the major news sources in our collective lives.
[17:47:40] <legoktm>	 citing Wikidata ends up falling under https://en.wikipedia.org/wiki/Wikipedia:Reliable_sources#User-generated_content which applies to Wikipedia itself
[19:01:58] <cdanis>	 the reliable sources page on enwiki is such an interesting lens through which to view all of modern society tbh
[19:07:15] <ebernhardson>	 Is it possible to perform automated actions based off alertmanager rules, such as restarting an instance? I'm thinking something like memory pressure + uptime > x days run a thing that will restart the instance
[19:07:53] <ebernhardson>	 i suppose it could run a cookbook that we pre-preared
[20:37:10] <volans|off>	 ebernhardson: yes and no. No strictly speaking right now, but enabling cookbooks runs "remotely" as an auto-remediation method is something in the TODO list. What can totally be done right now is the other way around, having either a daemon or systemd timer that runs a script (using spicerack) or a cookbook that checks either alertmanager or directly some metric on prometheus and then does what 
[20:37:16] <volans|off>	 you need.