scale-nodegroup.sh

Previous Next

Purpose

Update managed nodegroup scaling settings (min, max, desired) with safe validation.

Location

cloud/aws/eks/scale-nodegroup.sh

Preconditions

  • Required tools: bash, aws
  • Required permissions: eks:UpdateNodegroupConfig, eks:DescribeNodegroup, eks:DescribeUpdate
  • Required environment variables: none

Arguments

Flag Required Default Description
--cluster-name NAME Yes N/A Cluster name
--nodegroup-name NAME Yes N/A Nodegroup name
--min-size N Cond. current New min size
--max-size N Cond. current New max size
--desired-size N Cond. current New desired size
--wait / --no-wait No wait enabled Wait for update completion
--timeout SEC No 1800 Wait timeout
--poll-interval SEC No 20 Poll interval
--region REGION, --profile PROFILE No AWS defaults AWS context override
--dry-run No false Print planned commands

Scenarios

  • Happy path: scaling update completes successfully.
  • Common operational path: increase/decrease worker capacity based on workload demand.
  • Failure path: invalid requested bounds or update failure.
  • Recovery/rollback path: apply known-good scale values and retry update.

Usage

cloud/aws/eks/scale-nodegroup.sh --cluster-name prod-eks --nodegroup-name general --desired-size 6
cloud/aws/eks/scale-nodegroup.sh --cluster-name prod-eks --nodegroup-name general --min-size 3 --max-size 12 --desired-size 6
cloud/aws/eks/scale-nodegroup.sh --cluster-name dev-eks --nodegroup-name general --desired-size 2 --dry-run

Behavior

  • Main execution flow:
  • loads current scaling config
  • merges requested values
  • validates final bounds
  • submits update and optionally waits for success
  • Idempotency notes: convergent for same target scaling values.
  • Side effects: node count scaling and related workload scheduling impact.

Output

  • Standard output format: update ID on stdout; logs on stderr.
  • Exit codes:
  • 0 success
  • 2 invalid arguments/bound violations
  • non-zero on update failure or timeout

Failure Modes

  • Common errors and likely causes:
  • desired outside min/max
  • IAM denies update actions
  • nodegroup update conflicts/pending operations
  • Recovery and rollback steps:
  • correct scaling inputs and retry
  • wait for existing updates to complete
  • inspect update error details via describe-update

Security Notes

  • Secret handling: none.
  • Least-privilege requirements: update/read access limited to target nodegroups.
  • Audit/logging expectations: scaling changes should map to autoscaling policy/change records.

Testing

  • Unit tests:
  • bounds validation and update payload construction
  • Integration tests:
  • scale up/down in non-production nodegroups
  • Manual verification:
  • compare desired/current sizes after update completion
create-nodegroup.sh delete-nodegroup.sh