On 11/1/2019 10:22 AM, Matt Riedemann wrote:
On 11/1/2019 9:55 AM, Clark Boylan wrote:
OVH controls the disk IOPs that we get pretty aggressively as well. Possible it is an IO thing?
Yeah, so looking at the dstat output in that graph (thanks for pointing out that site, really nice) we basically have 0 I/O from 16:53 to 16:55, so uh, that's probably not good.
Dredging this back up after discussing with Clark in IRC today. I've posted a patch to disable c-bak and etcd3 in grenade jobs [1]. The hope is this will alleviate some pressure on OVH nodes where iops are restricted. We aren't even effectively doing anything with c-bak and etcd3 in grenade jobs so this should not be controversial. [1] https://review.opendev.org/#/c/700214/ -- Thanks, Matt