Wednesday, 2023-06-21

opendevreviewJames E. Blair proposed zuul/zuul-jobs master: Pin podman on jammy  https://review.opendev.org/c/zuul/zuul-jobs/+/88655201:51
*** dmellado170420 is now known as dmellado1704205:05
mnasiadkaclarkb: after our discussion in Vancouver it seems that we're getting a lot of NODE_FAILURE on arm nodes, example is current run of 886574 (5 node failures out of 8 jobs) - is there a way to make it better?08:02
fricklerdonate a stable arm64 cloud? *scnr*08:11
fricklerother than that, linaro is failing with "no valid host found" and osuosl with: No fixed IP addresses available for netw08:15
fricklerork: 2d0425cd-5898-434f-8669-f8ab2d5d0f72, not rescheduling.08:15
*** Tengu_ is now known as Tengu08:37
fungiwe can do a bit of looking on our end to check for stuff that can be deleted but wasn't automatically for some reason, though beyond that we'll need to reach out to the admins of those clouds to intervene on our behalf12:09
*** d34dh0r5- is now known as d34dh0r5312:12
*** mgoddard- is now known as mgoddard12:28
NeilHanlonfrickler, fungi: where are the arm nodes now?13:25
fungihalf in osuosl, half in an (equinix hosted?) colo run by linaro staff13:27
*** gthiemon1e is now known as gthiemonge16:00
mnasiadkafungi: understood, anything I can do to help (but doesn’t include donating ARM hardware)?18:51
fungimnasiadka: letting us know about the issues you run into is great. we're just stretched thin this week for a variety of reasons so i haven't had time to go looking any closer at it yet19:08
Clark[m]I think ianw or someone pointed out the linaro cert had expired19:18
Clark[m]During the summit19:18
fungioh, yes that could be it too but then i'd expect nodepool to throw connection errors instead of "no valid host found" coming back from nova19:23
Clark[m]Ah maybe that got fixed then or I misremembered 19:23
opendevreviewMerged zuul/zuul-jobs master: Pin podman on jammy  https://review.opendev.org/c/zuul/zuul-jobs/+/88655219:27
NeilHanloninfra-root: just fyi, we (Rocky) have been using the opendev etherpad for a discussion about RHEL source changes that were announced this morning and it's gotten some attention. Just thought I'd mention it as it has like ~140 viewers at the moment... which was not my intention lol19:49
fungiNeilHanlon: no worries, we've done far worse at ptgs. one trick is to distribute the read-only url for a pad in publications to cut down on accidental edits (we frequently see people with misbehaving auto-translation browser plugins mass translate the content into their language)19:58
fungialso be ready for defacements19:58
fungithere are no acls and we can't "lock down" a pad, so periodic exports are useful19:58
fungiwe can roll back to a specific earlier revision in a pinch, but it needs admin access to the rest api19:59
NeilHanlonoh, good call. thank you! I didn't know about the RO link20:01
fungialso, while it hasn't happened lately, we do occasionally see pads get "corrupted" in ways that are irrecoverable (though usually i can pull a json string object out of the database and get most of the content back, sans author colors and formatting)20:08
fungii don't recommend it for long-term storage/preservation of data20:08

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!