https://github.com/decayofmind/kube-better-node
A simple Descheduler-like program for preferredDuringSchedulingIgnoredDuringExecution nodeAffinity terms
https://github.com/decayofmind/kube-better-node
descheduler kubernetes soft workaround
Last synced: 5 months ago
JSON representation
A simple Descheduler-like program for preferredDuringSchedulingIgnoredDuringExecution nodeAffinity terms
- Host: GitHub
- URL: https://github.com/decayofmind/kube-better-node
- Owner: decayofmind
- License: mit
- Archived: true
- Created: 2021-07-27T22:24:09.000Z (almost 5 years ago)
- Default Branch: master
- Last Pushed: 2023-11-06T12:51:19.000Z (over 2 years ago)
- Last Synced: 2025-09-08T23:48:39.230Z (9 months ago)
- Topics: descheduler, kubernetes, soft, workaround
- Language: Go
- Homepage:
- Size: 162 KB
- Stars: 7
- Watchers: 2
- Forks: 2
- Open Issues: 7
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# ❗ ARCHIVED ❗
With [#1210](https://github.com/kubernetes-sigs/descheduler/pull/1210) being merged, [kubernetes-sigs/descheduler](https://github.com/kubernetes-sigs/descheduler) now respects `preferredDuringSchedulingIgnoredDuringExecution` term,
so there's no need to use and maintain this program
# kube-better-node
This small program does the only thing [kubernetes-sigs/descheduler](https://github.com/kubernetes-sigs/descheduler) can't do. It will evict Pods which could be potentially schedulled to a better Node, based on `preferredDuringSchedulingIgnoredDuringExecution` `nodeAffinity` terms.
### ⚠️ Disclaimer ⚠️
* I have no intention to replicate and follow all Descheduler policies and rules, but feel free to create a PR if something is missing.
* I do not need functionality provided by this program anymore, however I still see it useful for exploring the Golang world.
* Remember, it' a workaround and there's no guarantee that a Pod will be scheduled to a Node you expect it to be.
## Why?
There're at least two rotten unmerged PRs in the Descheduler repository:
* [#130](https://github.com/kubernetes-sigs/descheduler/pull/130) (`FindBetterPreferredNode` and `CalcPodPriorityScore` are inspired by it. Thanks [@tsu1980](https://github.com/tsu1980))
* [#129](https://github.com/kubernetes-sigs/descheduler/pull/129)
In fact, this functionallity will be never implemented in Descheduler.
As [@asnkh](https://github.com/asnkh) said in [#211](https://github.com/kubernetes-sigs/descheduler/issues/211#issuecomment-602026583):
> ...unless descheduler can make the same decision as kube-scheduler, it can cause this kind of ineffective pod evictions. I have no good idea to overcome this difficulty. Copying all scheduling policies in kube-scheduler to descheduler is not realistic.
## Use case
However, there're real world use-cases, where such dumb eviction can be usefull.
For example, when your project is small and there's a need to use some special Spot instance node type (such as Tesla `p2.xlarge`).
This type of instance can be taken back from you by AWS, but if your project can afford some performance degradation for a short period of time (untill there's a new node of `p2.xlarge` given), `kube-scheduler` will place Pods on some other Node available at the moment.
But when finally new Tesla `p2.xlarge` node will apper and join the cluster, there's nothing to schedule your Pods back to it.
Here **kube-better-node** can be usefull.
## Usage
```sh
❯ kube-better-node -h
Usage of kube-better-node:
-dry-run
Dry run
-tolerance int
Ignore certain weight difference
-v value
number for the log level verbosity
```
## Installation
```
helm repo add decayofmind https://decayofmind.github.io/charts/
helm install kube-better-node decayofmind/kube-better-node
```
## Links
* https://github.com/kubernetes-sigs/cluster-capacity