hi,
I'm looking into disk requirements for having swift in beta, ATM deployment-upload:/data/project/upload7 is ~73GB. I'd like to replicate data at least 2x in beta or ideally 3x, provisioning 3x xlarge instances seems unnecessary wasteful in terms of cpu/mem I think. What's the easiest way to get medium/large instances with additional (e.g 80-100G) local disk? I'm fine with provisioning xlarge instances of course if cpu/mem waste isn't a big concern
Description
Status | Subtype | Assigned | Task | |
---|---|---|---|---|
· · · | ||||
Resolved | • AlexMonk-WMF | T64835 Setup a Swift cluster on beta-cluster to match production | ||
Resolved | fgiunchedi | T123512 beta swift labs instances requirements | ||
· · · |
Event Timeline
We can probably make you special instances that have more disk and less CPU/RAM, but it's probably also ok to use xlarge since instance CPUs aren't really used when they aren't used.
/data/project/upload7 hasn't been cleaned up in ages.
- InstantCommons is enabled and thus lot of files are copied from production commons. We can probably nuke a lot of them
- We can probably nuke math, temp, thumb subdirectories
thanks @hashar ! I'd like to have some wiggle room just in case
anyways I don't seem to be able to add large/xlarge instances to deployment-prep ATM, quotas have been hit perhaps?
indeed instance limits (55) have been hit, https://wikitech.wikimedia.org/w/index.php?title=Special:NovaProject&action=displayquotas&projectname=deployment-prep can we bump that to +5 ?
@chasemp @yuvipanda @Andrew I'm assuming there's enough capacity to bump the limit and how to do that?
I just bumped up instance count limit for you guys! And \o/ to seeing this
happen, maybe we can finally get rid of NFS from deployment-prep!
{{done}}, required a #cores and memory bump too (following https://wikitech.wikimedia.org/wiki/OpenStack#Managing_project_quotas)