Gets rebalancing constraints for a cluster.

Gets rebalancing constraints for a cluster.
This helps the UI show what can/cannot be rebalanced before creating a plan.

Recent Requests
Log in to see full request history
TimeStatusUser Agent
Retrieving recent requests…
LoadingLoading…
Path Params
string
required

The ID of the Organization.

string
required

The ID of the cluster.

Query Params
boolean

Rebalance workloads that use local-path Persistent Volumes.
WARNING: THIS WILL RESULT IN DATA LOSS.

boolean

Pods spawned by Jobs or CronJobs will not prevent the Rebalancer from deleting a node on which they run.
WARNING: When true, pods spawned by Jobs or CronJobs will be terminated if the Rebalancer picks a node that runs them.
As such, they are likely to lose their progress.

boolean

Pods that are marked with "removal disabled" will not prevent the Rebalancer from deleting a node on which they run.
WARNING: When true, such pods will be evicted and disrupted.

boolean

Pods that don't have a controller (bare pods) will not prevent the Rebalancer from deleting a node on which they run.
WARNING: When true, such pods might not restart, since they have no controller to do it.

boolean

Ignore instance criteria constraints from node pools during rebalancing.
When true, removes requirements that constrain instance types by name patterns
(instance-category, instance-family, instance-generation, instance-size, instance-cpu, instance-memory).
This allows the rebalancer to select from a broader range of instance types.

boolean

Use aggressive eviction of NodeClaims which could block rebalancing plan execution
When true, NodeClaims and pods on them will be forcefully deleted after a default or specified
time interval has passed

string

drain_timeout_seconds specifies the maximum time allowed for draining nodes during aggressive eviction.
OPTIONAL: If not specified (0), nodes drain indefinitely (graceful drain).
When specified, valid range: 60-3600 seconds (1 minute to 1 hour).
(-- api-linter: core::0142::time-field-type=disabled
aip.dev/not-precedent: Using plain int64 seconds for simpler REST API client usage. --)

boolean

Ignore pods blocked by PodDisruptionBudgets with 0 disruptions allowed.
WARNING: May violate PDB guarantees during eviction.

boolean

Ignore NodePool disruption budgets that allow 0 disruptions.
When true, overrides zero-disruption NodePool budgets with the default 10% safety budget
so rebalancing can proceed.
WARNING: May violate NodePool disruption budget guarantees.

boolean

Nodes that are marked with "removal disabled" or "do-not-disrupt" annotations/labels
will not prevent the Rebalancer from deleting them.
WARNING: When true, such nodes will be disrupted despite their protection annotations.

boolean

When true, bypasses the Karpenter Consolidatable condition check during node
candidate selection. Nodes that Karpenter has not yet marked as consolidatable
(e.g. the ConsolidateAfter time window has not elapsed) will still be considered
for rebalancing.
WARNING: May disrupt nodes before Karpenter's own consolidation timing.

Responses

Language
Credentials
URL
LoadingLoading…
Response
Click Try It! to start a request and see the response here! Or choose an example:
application/json