655b4081f4
Migrate to klog v2
2020-06-05 17:22:26 +02:00
73a5cdf928
Address recent breaking changes in scheduler
...
The following things changed in scheduler and needed to be fixed:
* NodeInfo was moved to schedulerframework
* Some fields on NodeInfo are now exposed directly instead of via getters
* NodeInfo.Pods is now a list of *schedulerframework.PodInfo, not *apiv1.Pod
* SharedLister and NodeInfoLister were moved to schedulerframework
* PodLister was removed
2020-04-24 17:54:47 +02:00
8313e969c7
Add support for passing in custom ignore labels
2020-03-17 14:30:03 +11:00
5476125063
Use builder methods to create NodeInfoComparator functions
2020-03-17 13:51:15 +11:00
a7f3e54770
Add eks.amazonaws.com/nodegroup label to awsIgnoredLabels
2020-02-20 11:36:14 +01:00
d422aaaca6
UPSTREAM: <carry>: openshift: Add topology.kubernetes.io labels to be ignored when comparing similar node groups.
...
Without this, the autoscaler where using the lables in compareLabels and failing to match similar groups in different zones. Starting in kube 1.17 failure-domain.beta.kubernetes.io/* are deprecated in favour of topology.kubernetes.io/* https://kubernetes.io/docs/reference/kubernetes-api/labels-annotations-taints/#failure-domainbetakubernetesiozone
2020-02-19 18:15:11 +01:00
dde3341133
Raise maximum memory capacity difference.
...
AWS M5 instance types may differ in memory capacity by more than 128MB.
2019-10-25 17:18:08 -04:00
7f0a42b023
Add additional AWS labels.
...
Whitelist additional node labels for AWS CNI custom networking and
EC2 lifecycle.
Move AWS ignored node labels to AWS specific file.
2019-10-25 17:17:02 -04:00
7c9d6e3518
Do not assign return values to variables when not needed
2019-10-25 19:28:00 +08:00
dc1f19fc47
Merge pull request #2207 from viafoura/kops-node-similarity-fix-master
...
add kops instance group label to ignore list for similar node group identification.
2019-09-27 07:27:37 -07:00
e8b3c2a111
compare_nodegroups: Tolerate small differences in memory capacity
...
The current comparator expects memory capacity values to be identical.
However across AWS, Azure and GCP I quite often see very small
differences in capacity, typically 8-16Ki. When this occurs the
nodegroups are considered not equal when balancing is in effect which
is unfortunate because, in reality, they are identical.
This change will now tolerate a 128Ki difference before memory
capacity values are considered unequal.
2019-09-06 15:55:51 +01:00
75030ee2ec
Fix bug in balancing processor. Cluster Autoscaler was stopping scaling
...
up when there was a multizonal pool with number of nodes exceeding limit for one zone.
2019-07-29 09:28:20 +02:00
754412d7ea
also add similar label for eksctl to ignore list
...
Signed-off-by: Joe Hohertz <joe@viafoura.com>
2019-07-26 10:07:54 -04:00
1999d3b432
add kops instance group label to ignore list for similar node group identification.
...
Signed-off-by: Joe Hohertz <joe@viafoura.com>
2019-07-23 09:08:00 -04:00
89a09ccf00
Refactor the corresponding code.
2019-07-22 08:58:51 +08:00
f7c563ab06
Modify the code as the simple solution proposed by MaciekPytel.
2019-07-18 23:58:05 +08:00
622a838c2c
Modify nodal similarity rules.
2019-07-09 16:04:40 +08:00
34a4262ad8
Remove GKE specific node group comparator
...
Change-Id: I33131fec9b7972780cffde605a087cd2ad002752
2019-03-11 17:49:59 +01:00
2546d0d97c
Move leaderelection options to new packages
2019-02-21 13:45:46 +08:00
128729bae9
Move schedulercache to package nodeinfo
2019-02-21 12:41:08 +08:00
016bf7fc2c
Use k8s.io/klog instead github.com/golang/glog
2018-11-26 17:30:31 +01:00
01a56a8d73
Add GKE-specific NodeGroupSet processor
...
Also refactor Balancing processor a bit to make it easily extensible.
2018-10-25 18:50:17 +02:00
6f5e6aab6f
Move node group balancing to processor
...
The goal is to allow customization of this logic
for different use-case and cloudproviders.
2018-10-25 14:04:05 +02:00
Maciek Pytel