Closed praktiskt closed 10 months ago
Thank you for your pull request and welcome to the Trino community. We require contributors to sign our Contributor License Agreement, and we don't seem to have you on file. Continue to work with us on the review and improvements in this PR, and submit the signed CLA to cla@trino.io. Photos, scans, or digitally-signed PDF files are all suitable. Processing may take a few days. The CLA needs to be on file before we merge your changes. For more information, see https://github.com/trinodb/cla
Signed and sent the CLA.
Thank you for your pull request and welcome to the Trino community. We require contributors to sign our Contributor License Agreement, and we don't seem to have you on file. Continue to work with us on the review and improvements in this PR, and submit the signed CLA to cla@trino.io. Photos, scans, or digitally-signed PDF files are all suitable. Processing may take a few days. The CLA needs to be on file before we merge your changes. For more information, see https://github.com/trinodb/cla
CLA accepted by Trino.
@cla-bot check
The cla-bot has been summoned, and re-checked this pull request!
@mosabua @martint is there anything more needed before this can be merged?
Since #120 was merged after I created this (and this PR seems to have been abandoned by Trino maintainers), I'm going to go ahead and close it.
Adds:
apiVersion
fortemplates/autoscaling.yaml
- you can now setautoscaling/v2
if your cluster supports it.server.autoscaling.extraSpec
fortemplates/autoscaling.yaml
- pass whatever you want tospec
, meaning you can configure advanced behavior and metrics.Rationale: While the current implementation of
server.autoscaling
works for basic cases, it's often desired to expand and customize the behavior of auto scaling further. Some snags I've run into primarily revolves around workers scaling down too fast, causing tasks to fail. Allowing for templating the ~entire~ a good chunk of thespec
means you can configure things likebehavior
(slow down up-/downscaling,stabilizationWindowSeconds
) and moremetrics
(e.g. scale on memory limits).Actual autoscaling behavior is unchanged from previous version to this. Happy to bump this if we agree that
autoscaling/v2
is the default (meaning we can also set more sensible defaults forextraSpec
).