[04:55:45] <marostegui>	 I am going to switchover s5 primary db master in 5 minutes
[05:01:34] <marostegui>	 Everything is done
[07:10:08] <XioNoX>	 I'm going to failover VRRP from cr1-eqiad to cr2-eqiad in prevision of the PDU work of T226782, no impact expected
[07:10:09] <stashbot>	 T226782: a1-eqiad pdu refresh (Tuesday 10/15 @11am UTC) - https://phabricator.wikimedia.org/T226782
[10:40:49] <vgutierrez>	 https://tls.ulfheim.net/ TLSv1.2 and TLSv1.3 https://tls13.ulfheim.net/ for humans :)
[10:48:47] <effie>	 neat
[11:36:47] <_joe_>	 jbond42: FYI I had to change the ca server in cergen's configuration; when (if) you switch back to eqiad, please do change that as well
[11:36:59] <_joe_>	 it's in the private repo, you can even just revert my commit
[11:38:54] <jbond42>	 ack thanks i just reverted so will take a look now
[11:41:16] <jbond42>	 done
[11:52:24] <jbond42>	 lunch and errand
[11:53:28] <_joe_>	 jbond42: so we're back with puppetmaster1001 being the ca, correct?
[12:51:56] <arturo>	 jbond42: do you remember the puppet agent cmdline to check the value of a hiera key on a given server?
[12:53:45] <arturo>	 nevermind
[17:23:10] <mutante>	 when i look at LVS config, (hieradata/common/lvs/configuration.yaml) there are references to "IP blocks". (&ip_block057, *ip_block001 *ip_block002 etc..). Where are those defined? It does not seem to be the DNS repo ?
[17:25:29] <mutante>	 i want to add a new service and most things can be copy/pasted or the change is relatively obvious, but i'm sure i can't just use the same block as in "ip: *ip_block011"
[17:29:47] <mutante>	 oh. think i got it. all in that same file
[18:55:26] <_joe_>	 yes
[18:55:40] <_joe_>	 also, you will use the same IPs as parsoid
[18:58:25] <mutante>	 oh, then it was actually right in the previous patch set. ok
[19:11:43] <_joe_>	 yes, we will use same IP but different port - I would suggest using 443 and https from the get go
[19:12:42] <_joe_>	 so the idea is
[19:12:52] <_joe_>	 you first add the stuff to conftool-data
[19:13:15] <_joe_>	 then you set all the nodes where we still didn't install MediaWiki to pooled=inactive
[19:13:34] <_joe_>	 so they're not visible anywhere
[19:14:12] <_joe_>	 then you add parsoid/parsoid-php to the scap::dsh::groups definition so that scap syncs to the servers where we installed mediawiki
[19:14:28] <_joe_>	 to the mediawiki-installation dsh group I mean
[19:14:37] <_joe_>	 LVS can happen afterwards
[19:15:24] <mutante>	 ok! cool. i'll get the patches in order
[19:15:56] <_joe_>	 the most important part is getting the code to be in sync
[19:16:21] <_joe_>	 so that the parsing team can run their tests