Skip to content

upgrades.md

Latest commit

 

History

History
91 lines (64 loc) · 2.17 KB

upgrades.md

File metadata and controls

91 lines (64 loc) · 2.17 KB

Instance upgrades and maintenance

Check available upgrade versions

cloudamqp instance upgrade-versions --id 1234
cloudamqp instance nodes versions --id 1234

Upgrade paths

Upgrade everything (recommended)

Upgrades both Erlang and RabbitMQ/LavinMQ to the latest compatible versions:

cloudamqp instance upgrade-all --id 1234

Upgrade individually

# erlang first, then RabbitMQ
cloudamqp instance upgrade-erlang --id 1234
cloudamqp instance upgrade-rabbitmq --id 1234 --version=3.13.0

Always upgrade Erlang before RabbitMQ when doing both separately.

Restart operations

# restart RabbitMQ process (rolling for HA clusters)
cloudamqp instance restart-rabbitmq --id 1234

# restart specific nodes only
cloudamqp instance restart-rabbitmq --id 1234 --nodes=node1,node2

# full cluster restart
cloudamqp instance restart-cluster --id 1234

# restart management interface only
cloudamqp instance restart-management --id 1234

Start, stop, and reboot

cloudamqp instance stop --id 1234
cloudamqp instance start --id 1234
cloudamqp instance reboot --id 1234

# cluster-level
cloudamqp instance stop-cluster --id 1234
cloudamqp instance start-cluster --id 1234

Disk resize

# resize disk (may require downtime)
cloudamqp instance resize-disk --id 1234 --disk-size=100 --allow-downtime

Maintenance workflow

A typical maintenance sequence:

# 1. check current state
cloudamqp instance get --id 1234
cloudamqp instance nodes list --id 1234

# 2. check available upgrades
cloudamqp instance upgrade-versions --id 1234

# 3. perform upgrade
cloudamqp instance upgrade-all --id 1234

# 4. verify after upgrade
cloudamqp instance nodes list --id 1234
cloudamqp instance nodes versions --id 1234

Important notes

  • Upgrade operations are async: they return immediately but run in the background.
  • Poll instance get --id <id> to check when the instance is ready again.
  • Plan changes via instance update --id <id> --plan=<plan> may cause brief downtime.
  • restart-rabbitmq does a rolling restart on HA clusters (minimal disruption).
  • restart-cluster restarts all nodes simultaneously (causes downtime).