[10:44:58] <_joe_>	 James_F: did I say we can drop hhvm completely?
[10:45:15] <_joe_>	 I must have either misunderstood or brainfarted if i did
[10:45:42] <_joe_>	 I was aware of wikitech and also i want to remove hhvm from at least mwmaint
[10:46:05] <_joe_>	 so if I gave you the green light, I either misunderstood or I did have a lapse of reason
[10:47:01] <_joe_>	 A more savvy me would've said "next week if possible" :P
[12:32:35] <mark>	 where can we not yet remove hhvm?
[12:56:30] <_joe_>	 from mediawiki, given it's still used by wikitech and it's still installed everywhere
[12:57:04] <_joe_>	 I think the most important part is removing it from the maintenance servers
[12:59:10] <mark>	 why is it still used by wikitech?
[12:59:57] <_joe_>	 because wikitech has its own puppet classes, separate from the main clusters, and we never got to it
[13:00:07] <_joe_>	 it's not a technical challenge
[13:03:35] <mark>	 how much work would it be to remove it?
[13:09:17] <_joe_>	 remove HHVM from the mwmaint servers? not much
[13:10:06] <_joe_>	 if you mean how much work it is to move wikitech to php7, I don't think it's a lot of work either.
[13:11:39] <_joe_>	 andrewbogott: lmk if you need help with that.
[13:14:07] <andrewbogott>	 thanks!  I think bd.808 and bs.torm_ were going to work on that today
[13:17:12] <_joe_>	 oh ok sorry, I got that backwards
[13:39:32] <mark>	 cool
[13:49:28] <James_F>	 _joe_: As I said, we can hold back the train from wikitech for a while. But yes, whoops.
[14:10:43] <effie>	 James_F: I think that makes sense, wmcs will be finish this up sooner or later
[18:32:49] <chaomodus>	 question, I'd liek to make netbox active/active since it works fine that way, but dns ci complains about multiple addresses for a name
[18:32:59] <chaomodus>	 I gather this is not the normal way to do this, then
[18:36:32] <godog>	 chaomodus: yeah it'll be through dns discovery records that feed from conftool-data
[18:36:45] <chaomodus>	 oh dear
[18:36:47] <chaomodus>	 okay
[18:38:16] <chaomodus>	 thanks for the pointer, i found the wikyitech article on it
[18:39:08] <chaomodus>	 except this doesn't seem to be for wikimedia.org addresses?
[18:40:01] <godog>	 chaomodus: oh, my bad I defaulted to internal discovery.wmnet
[18:41:20] <godog>	 trying to think if we have sth like that already, i.e. that doesn't go through cache
[18:42:12] <mutante>	 is netbox not supposed to be behind caching servers because it is considered like a monitoring server that should not rely on it?
[18:42:23] <chaomodus>	 Yah
[18:42:31] <chaomodus>	 also a source of configuration truth going forward
[18:42:39] <mutante>	 gotcha.. or i would have said stuff about config for ATS.. nvm then
[18:42:49] <chaomodus>	 yah preferably not behind balancers
[18:44:23] <godog>	 geodns it is then, in other words the DYNA configuration in dns and friends, with its own pool
[18:45:35] <chaomodus>	 mh yah
[18:45:37] <chaomodus>	 okay
[18:47:14] <chaomodus>	 so that's like make a new key in geo-resources
[18:47:17] <chaomodus>	 with the addrs
[18:47:28] <chaomodus>	 and make it geoip!netbox-addrs or whatever
[18:47:33] <chaomodus>	 it the dns
[18:47:40] <chaomodus>	 that's easier than expected.
[18:54:42] <godog>	 afaik that's what there is to it yes, traffic folks can confirm
[18:57:39] <chaomodus>	 cool thanks :)
[18:57:45] <chaomodus>	 what even is this format though
[19:03:30] <chaomodus>	 who should review this now?
[19:03:49] <chaomodus>	 idk, anyone qualified to this weird part of dns: https://gerrit.wikimedia.org/r/c/operations/dns/+/541602
[19:07:55] <godog>	 traffic / brandon would be it, also the related task in commit will help with context
[19:16:05] <chaomodus>	 Yes