Considerations for grid node decommission
PDF of this doc site
- Get started
Install and maintain appliance hardware
SG100 and SG1000 services appliances
- Prepare for installation (SG100 and SG1000)
SG6000 storage appliances
- Prepare for installation (SG6000)
- Configure hardware (SG6000)
SG5700 storage appliances
- Prepare for installation (SG5700)
- Configure hardware (SG5700)
SG5600 storage appliances
- Prepare for installation (SG5600)
- Configure hardware (SG5600)
- SG100 and SG1000 services appliances
Install and upgrade software
- Upgrade StorageGRID software
- Install Red Hat Enterprise Linux or CentOS
- Install Ubuntu or Debian
Perform system administration
- Manage security settings
- Manage Admin Nodes
- Manage Archive Nodes
Manage objects with ILM
- ILM and object lifecycle
- Create storage grades, storage pools, EC profiles, and regions
- Administer StorageGRID
- Use a tenant account
- S3 REST API supported operations and limitations
Monitor and maintain StorageGRID
Monitor and troubleshoot
- Troubleshoot a StorageGRID system
- Expand your grid
Recover and maintain
Grid node recovery procedures
- Recover from Storage Node failures
- Recover from Admin Node failures
- All grid node types: Replace Linux node
- Grid node decommission
- Network maintenance procedures
- Grid node procedures
- Grid node recovery procedures
Review audit logs
- Audit messages and the object lifecycle
- Monitor and troubleshoot
Before you start this procedure to decommission one or more nodes, you must understand the implications of removing each type of node. Upon the successful decommissioning of a node, its services will be disabled and the node will be automatically shut down.
You cannot decommission a node if doing so will leave StorageGRID in an invalid state. The following rules are enforced:
You cannot decommission the primary Admin Node.
You cannot decommission Archive Nodes.
You cannot decommission an Admin Node or a Gateway Node if one of its network interfaces is part of a high availability (HA) group.
You cannot decommission a Storage Node if its removal would affect the ADC quorum.
You cannot decommission a Storage Node if it is required for the active ILM policy.
You should not decommission more than 10 Storage Nodes in a single Decommission Node procedure.
You cannot decommission a connected node if your grid includes any disconnected nodes (nodes whose health is Unknown or Administratively Down). You must decommission or recover the disconnected nodes first.
If your grid contains multiple disconnected nodes, the software requires you to decommission them all at the same time, which increases the potential for unexpected results.
If a disconnected node cannot be removed (for example, a Storage Node that is required for the ADC quorum), no other disconnected node can be removed.
If you want to replace an older appliance with a newer appliance, consider cloning the appliance node instead of decommissioning the old node and adding the new node in an expansion.
|Do not remove a grid node’s virtual machine or other resources until instructed to do so in decommission procedures.|