Starting with descheduler release v0.10.0 container images are available in the official k8s container registry.
| Descheduler Version | Container Image | Architectures |
|---|---|---|
| v0.21.0 | k8s.gcr.io/descheduler/descheduler:v0.21.0 | AMD64 ARM64 ARMv7 |
| v0.20.0 | k8s.gcr.io/descheduler/descheduler:v0.20.0 | AMD64 ARM64 |
| v0.19.0 | k8s.gcr.io/descheduler/descheduler:v0.19.0 | AMD64 |
| v0.18.0 | k8s.gcr.io/descheduler/descheduler:v0.18.0 | AMD64 |
| v0.10.0 | k8s.gcr.io/descheduler/descheduler:v0.10.0 | AMD64 |
Note that multi-arch container images cannot be pulled by kind from a registry. Therefore starting with descheduler release v0.20.0 use the below process to download the official descheduler image into a kind cluster.
kind create cluster
docker pull k8s.gcr.io/descheduler/descheduler:v0.20.0
kind load docker-image k8s.gcr.io/descheduler/descheduler:v0.20.0
The examples directory has descheduler policy configuration examples.
The descheduler has many CLI options that can be used to override its default behavior.
descheduler --help
The descheduler evicts pods which may be bound to less desired nodes
Usage:
descheduler [flags]
descheduler [command]
Available Commands:
help Help about any command
version Version of descheduler
Flags:
--add-dir-header If true, adds the file directory to the header of the log messages
--alsologtostderr log to standard error as well as files
--descheduling-interval duration Time interval between two consecutive descheduler executions. Setting this value instructs the descheduler to run in a continuous loop at the interval specified.
--dry-run execute descheduler in dry run mode.
--evict-local-storage-pods DEPRECATED: enables evicting pods using local storage by descheduler
-h, --help help for descheduler
--kubeconfig string File with kube configuration.
--log-backtrace-at traceLocation when logging hits line file:N, emit a stack trace (default :0)
--log-dir string If non-empty, write log files in this directory
--log-file string If non-empty, use this log file
--log-file-max-size uint Defines the maximum size a log file can grow to. Unit is megabytes. If the value is 0, the maximum file size is unlimited. (default 1800)
--log-flush-frequency duration Maximum number of seconds between log flushes (default 5s)
--logtostderr log to standard error instead of files (default true)
--max-pods-to-evict-per-node int DEPRECATED: limits the maximum number of pods to be evicted per node by descheduler
--node-selector string DEPRECATED: selector (label query) to filter on, supports '=', '==', and '!='.(e.g. -l key1=value1,key2=value2)
--policy-config-file string File with descheduler policy configuration.
--skip-headers If true, avoid header prefixes in the log messages
--skip-log-headers If true, avoid headers when opening log files
--stderrthreshold severity logs at or above this threshold go to stderr (default 2)
-v, --v Level number for the log level verbosity
--vmodule moduleSpec comma-separated list of pattern=N settings for file-filtered logging
Use "descheduler [command] --help" for more information about a command.
This section contains descriptions of real world production use cases.
When initially migrating applications from a static virtual machine infrastructure to a cloud native k8s infrastructure there can be a tendency to treat application pods like static virtual machines. One approach to help prevent developers and operators from treating pods like virtual machines is to ensure that pods only run for a fixed amount of time.
The PodLifeTime strategy can be used to ensure that old pods are evicted. It is recommended to create a
pod disruption budget for each
application to ensure application availability.
descheduler -v=3 --evict-local-storage-pods --policy-config-file=pod-life-time.yml
This policy configuration file ensures that pods created more than 7 days ago are evicted.
---
apiVersion: "descheduler/v1alpha1"
kind: "DeschedulerPolicy"
strategies:
"PodLifeTime":
enabled: true
params:
maxPodLifeTimeSeconds: 604800 # pods run for a maximum of 7 days
If your cluster has been running for a long period of time, you may find that the resource utilization is not very
balanced. The following two strategies can be used to rebalance your cluster based on cpu, memory
or number of pods.
Using LowNodeUtilization, descheduler will rebalance the cluster based on memory by evicting pods
from nodes with memory utilization over 70% to nodes with memory utilization below 20%.
apiVersion: "descheduler/v1alpha1"
kind: "DeschedulerPolicy"
strategies:
"LowNodeUtilization":
enabled: true
params:
nodeResourceUtilizationThresholds:
thresholds:
"memory": 20
targetThresholds:
"memory": 70
Using HighNodeUtilization, descheduler will rebalance the cluster based on memory by evicting pods
from nodes with memory utilization lower than 20%. This should be used along with scheduler strategy MostRequestedPriority.
The evicted pods will be compacted into minimal set of nodes.
apiVersion: "descheduler/v1alpha1"
kind: "DeschedulerPolicy"
strategies:
"HighNodeUtilization":
enabled: true
params:
nodeResourceUtilizationThresholds:
thresholds:
"memory": 20
Descheduler's RemovePodsViolatingNodeTaints strategy can be combined with
Node Problem Detector and
Cluster Autoscaler to automatically remove
Nodes which have problems. Node Problem Detector can detect specific Node problems and taint any Nodes which have those
problems. The Descheduler will then deschedule workloads from those Nodes. Finally, if the descheduled Node's resource
allocation falls below the Cluster Autoscaler's scale down threshold, the Node will become a scale down candidate
and can be removed by Cluster Autoscaler. These three components form an autohealing cycle for Node problems.