[07:46:40] fyi, looks like netbox exports some drmrs addresses to esams files? https://phabricator.wikimedia.org/rONED05120666db24176e495d9f8dea1ae90bda980110 [07:55:24] moritzm: good catch, I'll check with volans|off on Monday [08:10:55] FWIW I was looking through the code that generates the DNS names, and comparing to the data in Netbox. [08:11:15] Can't spot where the issue is, the data for DRMRS in Netbox doesn't look to have any missing or typos that I can see. [08:11:36] I'm sure volan.s will quickly spot the issue of course :) [08:12:30] yeah, we were discussing it... so much for a |off [08:12:41] heh yeah. [08:12:57] he was helping me with stuff a short time ago, and when we were finished change his nick to |off [08:13:12] I was like damn this isn't important enough if it's you day off ffs! [08:13:20] the topmost /32 is assigned to esams [08:13:49] but if we set it to no-site (as it should now) everything goes to -global [08:13:54] which one? [08:14:02] https://netbox.wikimedia.org/ipam/prefixes/402/ [08:14:26] so the code needs to be changed so it browse the tree from more specific to less specific and stops on the first hit [08:14:36] as all the sub-prefixes have the correct site set [08:14:42] ah ok gotcha. [08:15:05] yeah I went up from the IP in the parent prefixes and seen the parent was in drmrs, stopped there. [08:15:41] It would have taken me quite some time to figure out what's going on :) [08:18:05] wasn't jumping out at me it'd use least specific from the code [08:18:09] think it's maybe this "min" [08:18:13] if matching_prefixes: [08:18:13] prefix_key = min(matching_prefixes, key=attrgetter('prefixlen')) [08:18:57] exactly [08:23:34] Also I guess the /32 probably shouldn't be attached to any site? [08:24:41] topranks: yep, needs to be updated after the script is fixed, otherwise it will move everything to -global [08:28:29] 10netbox, 10Infrastructure-Foundations: Netbox/DNS, browse parents prefixes to set site - https://phabricator.wikimedia.org/T294082 (10ayounsi) [08:28:36] opened https://phabricator.wikimedia.org/T294082 [08:29:08] and I meant majavah above, not moritzm :) [08:34:28] https://gerrit.wikimedia.org/r/732919 [08:34:55] untested, but in theory should work [08:40:36] great, thanks! [09:16:10] majavah: I think that will fix it in this case, but unless I'm reading it wrong won't that just find the longest parent prefix the IP is part of? [09:16:56] i.e. it won't recurse up through the parent prefixes from most specific -> least specific until it finds one with a site attribute? [09:17:00] topranks: it finds the most specific prefix that IP is a part of and has a site attached to it [09:17:19] sry... [09:17:20] if self.netbox.prefixes[prefix].site [09:17:20] line 474 filters for only prefixes with sites attached [09:17:23] I think [09:17:40] my bad yeah... time for coffee I think :) [09:19:44] nice one yeah it makes sense to me [09:25:41] 10netops, 10DC-Ops, 10Infrastructure-Foundations, 10SRE, and 2 others: Q1:(Need By: TBD) rack/setup/install cloudswift100[12] - https://phabricator.wikimedia.org/T289882 (10ayounsi) Thanks! > I take it the main concern here is allocating a public IPv4 address, which is a scarce resource, no? That's one of... [10:26:32] hahaha just noticed that https://phabricator.wikimedia.org/T293588 starts with "Houston, we have a problem." [10:37:25] 10netops, 10DC-Ops, 10Infrastructure-Foundations, 10SRE, and 2 others: Q1:(Need By: TBD) rack/setup/install cloudswift100[12] - https://phabricator.wikimedia.org/T289882 (10aborrero) a:05aborrero→03ayounsi >>! In T289882#7450242, @ayounsi wrote: > Which means increasing our attack surface as well as SR... [11:56:25] 10Puppet, 10Infrastructure-Foundations, 10observability, 10Patch-For-Review, 10User-jbond: Add additional prometheus metrics to puppet runs - https://phabricator.wikimedia.org/T283585 (10fgiunchedi) @jbond IIRC for this we went the logstash way, anything else to be done and/or missing ? [11:58:13] 10Puppet, 10Infrastructure-Foundations, 10observability, 10Patch-For-Review, 10User-jbond: Add additional prometheus metrics to puppet runs - https://phabricator.wikimedia.org/T283585 (10jbond) 05Open→03Resolved a:03jbond Thats correct all though it still a work in progress, however this one i thin... [13:23:22] 10netops, 10Infrastructure-Foundations: Validate EVPN/VXLAN configuration for Juniper QFX Platform - https://phabricator.wikimedia.org/T294115 (10cmooney) p:05Triage→03Medium [13:46:07] Heads up: it's a public holiday in Ireland on Monday, so I'm not sure what that means for our team meeting. [13:49:15] jobo: any thoughts? [13:50:26] I'm also off on Monday (and the rest of the week) [13:53:28] oh nice... enjoy :) [14:04:27] I leave it up to the rest of the team, you can either meet or have it async. Etherpad is prepared :) https://etherpad.wikimedia.org/p/SRE-Foundations-2021-10-25 [15:09:08] I'll be in a train to Marseille on Monday during our team meeting [17:06:54] 10Packaging, 10Infrastructure-Foundations, 10SRE, 10serviceops: package requirements for upgrading deployment_servers to buster - https://phabricator.wikimedia.org/T242480 (10Majavah) 05Open→03Resolved This was done at some point. [18:14:58] 10Packaging, 10Infrastructure-Foundations, 10SRE, 10serviceops: package requirements for upgrading deployment_servers to buster - https://phabricator.wikimedia.org/T242480 (10Dzahn)