The default strategy was RollingUpdate with a maxSurge of 25% and a
maxUnavailble of 25%. This could cause the cluster autoscaler to add
more nodes to the cluster any time we update the overprovisioner
deployment.
Changing the strategy to Recreate seems sensible as this would keep the
same number of nodes on the cluster during updates.
Signed-off-by: Cezar Sa Espinola <cezarsa@gmail.com>
* PE-12185 Added labels and annotations to the deployments
Signed-off-by: Hingst, Arne-Kristian <arne@hingst.it>
* Fixes linter errors about to many empty lines
Signed-off-by: Hingst, Arne-Kristian <arne@hingst.it>
* Another blank line linter error
Signed-off-by: Hingst, Arne-Kristian <arne@hingst.it>
* [stable/cluster-overprovisioner] Add fleetName label
This label (`app.kubernetes.io/fleetName`) allows to match on a specific
overprovisioning deployment with matchExpressions/matchLabels in
affinity rules. The fleetName label will be unique per deployment fleet.
Signed-off-by: Koen van Ingen <koenvaningen@gmail.com>
* [stable/cluster-overprovisioner] Change fleetName label to cluster-overprovisioner-name
Signed-off-by: Koen van Ingen <koenvaningen@gmail.com>
* Update configuration for cluster-autoscaler 1.12
cluster-autoscaler >=1.12 uses a compatible default priority cutoff and requires no configuration changes.
Signed-off-by: Tim Hobbs <timothy.hobbs@ic-consult.com>
* Bump version
Signed-off-by: Tim Hobbs <timothy.hobbs@ic-consult.com>
* # This is a combination of 2 commits.
# This is the 1st commit message:
Initial commit
Signed-off-by: Max Williams <max.williams@deliveryhero.com>
# This is the commit message #2:
updating sources
* Commit Sign-off
Signed-off-by: Miguel Angel Mingorance Fernandez <miguel.mingorance@deliveryhero.com>
* Updating README table
Signed-off-by: Miguel Angel Mingorance Fernandez <miguel.mingorance@deliveryhero.com>
Matt Farina
Miguel MingoranceandMax Williams
Peter Wilson
Marcos Estévez
Cezar Sá Espinola
Hingst, Arne-Kristian
Koen van Ingen
dabde
Tim Hobbs
Taehyun Kim