09:40:37 <alane> I have discovered that we have lots of conflicting services_bom/eom in our MODB databases, does anyone know if this is normal beahaviour? I found this useful command that will show all conflicting documents in a database [https://docs.couchdb.org/en/stable/replication/conflicts.html#finding-conflicted-documents-with-mango], it needs couchdb rather than bigcouch but we've already upgraded to couchdb. I used this to write a 
09:40:37 <alane> quick script to audit all our MODBs and found that just about every one has conflicts on these 2 documents. When I examine the revisions it seems that different nodes in the zone (we have 3 apps servers) tried to write the same document at the same time, thats wierd I thought, so I took a look at kt_services_rollover.erl and it seems to me that there is no mechanism to prevent all the nodes trying to do the rollup at precisely the 
09:40:37 <alane> same time [midnight UTC on 1st of every month], any thoughts anyone?
15:48:30 <mc_> alane: yeah, there are a couple tasks like that
15:48:34 <mc_> needing more coordination
15:48:58 <mc_> in 5.x it is shifting to using the Raft stuff to find a "worker" within the cluster of `task` apps
15:49:02 <mc_> to do a job like that
16:57:49 <mentax> mc_ are there will be any changes in billing portion (braintree) in v5?
17:11:53 <mc_> mentax: not that i'm aware of (at least, i haven't seen PRs related to it)
17:12:13 <mc_> but other teams may have snuck some in :)
17:30:29 <bongfrog> 2 weeks