Increase cpu limits for postgres#636
Conversation
This reverts commit 453ef72. If we don't set cpu limits then default limits appear (500m). We need more than this in different pods, thus I keep the limits and increase it even for postgres.
Today we were above 1 cpu usage. Let put here the max we can. I also slightly increased valkey and pushgateway since sometimes we are close to the limit (looking at metrics).
The host key is not saved in the template anymore (for security reasons). Skip checking differences in this key between template and local file.
|
Build succeeded. ✔️ pre-commit SUCCESS in 1m 27s |
When conditionalizing the templates for volume claim, do not create even an empty object when it's not defined. Better to avoid it altogether than nesting something by mistake. Signed-off-by: Matej Focko <mfocko@redhat.com>
URL contains colon for the protocol, quote it to prevent YAML parsing issues. Signed-off-by: Matej Focko <mfocko@redhat.com>
|
Build succeeded. ✔️ pre-commit SUCCESS in 1m 27s |
lbarcziova
left a comment
lbarcziova
left a comment
There was a problem hiding this comment.
we could increase the CPU limit to 2 (the max allowed by cluster iiuc) also for other pods if needed, but from the metrics it looks like those are usually not problematic
yes, I discussed the same with @mfocko and we agreed that we should be ok with this limits for now 🤞🏻. |
|
Build succeeded. ✔️ pre-commit SUCCESS in 1m 26s |
majamassarini
force-pushed
the
increase-cpu-limits-for-postgres
branch
from
e120365 to
972d20a
Compare
|
Build succeeded. ✔️ pre-commit SUCCESS in 1m 28s |
|
Build succeeded (gate pipeline). ✔️ pre-commit SUCCESS in 1m 33s |
3 participants
If we don't set cpu limits they are being set by default (500m) but this value is too low for some of our pods.
Keep cpu limits.
Increase postgres cpu limit.