[10:25:11] <jbond>	 godog: fyi im reverting the get_clusters change, puppet runs are failing to complete on prometheus hosts.  illl try re-wrting wmflib::get_clusters in ruby
[10:26:08] <godog>	 jbond: ah, on what hosts? I'd like to take a look
[10:26:24] <jbond>	 godog: i assume all but i have been testing on prometheus1005
[10:26:31] <godog>	 ack
[10:27:09] <godog>	 jbond: but sure revert lgtm
[10:27:12] <jbond>	 it took about 15 minutes before it finnaly died
[10:27:33] <jbond>	 happy to hold of on the revert if you want to look just let me know when to merge
[10:29:00] <godog>	 jbond: please go ahead! thank you, it is clear that I should update the pontoon stack since I didn't get any failures when testing
[10:34:57] <jbond>	 godog: i think its just the difference in the number of hosts
[10:35:03] <jbond>	 pcc was also fine
[10:35:30] <godog>	 the number of hosts in the puppetdb query result ?
[10:35:39] <jbond>	 yes i think so
[10:35:48] <godog>	 ack
[10:35:50] <jbond>	 s/hosts/resources/
[10:37:06] <godog>	 ah ok got it, shouldn't pcc have failed too in that case? i.e. the puppetdb for pcc should have most/all resources loaded  IIRC?
[10:37:21] <jbond>	 godog: yes i was just thinking about that myself
[10:37:37] <jbond>	 if anything it has more as it also has cloud stuff
[10:38:53] <godog>	 that too, true