|mediawiki/services/ocg-collection||master||+8 -0||Add scap3 config|
|operations/puppet||production||+9 -6||Deploy ocg with scap3|
|Resolved||MoritzMuehlenhoff||T164780 Sunset our use of Salt|
|Resolved||thcipriani||T129290 [keyresult] Migrate remaining trebuchet deployed services|
|Declined||None||T129142 Deploy ocg with scap3|
See T150871: [EPIC] (Proposal) Replicate core OCG features and sunset OCG service. Original timeline / agreement with Reading was to sunset OCG this month. This has not happened yet, but there is some movement. Both ops & services clearly want this to happen ASAP, so that we can stop spending time on it for things like the Node security upgrade or Scap migration.
There is a working group (that I am on, that we just made on Monday, of tech-mgt people) to work through what it means and how to sunset things here. I hope we have something useful in this case by the end of the month-ish so $whoever_is_tasked_with_it can have time to migrate or remove by the end of Sept. That all depends on what the sunsetting process looks like (it might take longer than a month in some cases, is my guess).
The original deadline was 2017-08-01, but that hasn't happened yet. Sunsetting it by EOQ is possible, but it's a stretch IMHO.
The migration to Scap3 shouldn't take more than a day and should be straightforward AFAICT, so if you (RelEng) still want to do it, you should coordinate with @cscott and OpsEns.
Salt (on which Trebuchet is build on) is being removed by end of September. in TechOps we'll need a few days to fully clean out Salt from production.
Is the plan to still move ocg to scap3 despite it's remaining two weeks of lifetime or is it acceptable to dismantle the deployment basis of ocg starting the week of the 25th? If no further ocg deployments are planned, we could just as well rsync the fix manually in case of an unexpected ocg emergency deployment, the ocg* cluster only consists of three hosts.
I'm declining this task as ocg is meant to be replaced by the end of the quarter and (afaik) no further deployments are scheduled. I've created patches to migrate ocg should the need arise to do a deploy, but I don't think this should block the removal of salt.