Maintenance Schedule
A maintenance schedule shows which tasks a cluster administrator should perform daily, weekly, monthly, and quarterly, as well as initial setup tasks. This sample schedule is offered as a template that you can customize to suit your needs.
Initial Setup | Daily | Weekly | Monthly | Quarterly |
---|---|---|---|---|
Perform hardware checks | Check alarms |
Check logs (cluster and job logs) |
Clean up logs | Audit storage use |
Check Prerequisites | Onboard new users | Check for zombie jobs | Reevaluate node and volume topology | Perform security audit |
Verify Installation | Offboard users | Perform simple performance test | Check performance (jobs, I/O) | Upgrade MapR core, HBase, Hive, HCatalog |
High Availability planning | Check cluster health | Verify snapshots | Verify mirrors | |
Set up VIPs for NFS | Perform hardware maintenance (disks) | Remove old snapshots | Clean up data | |
Benchmark and tuning | Check TaskTracker blacklist | Failbacks to restore service layout | MFS balancing | |
Set up node and volume topologies | HBase balancing | |||
Capacity planning | Upgrade ecosystem components (other than HBase, Hive, and HCatalog, which are done quarterly) | |||
Set up user permissions (ACLs, directory permissions, users, home directories) |
||||
Set up quotas | ||||
Set up compression | ||||
Configure cluster queues for MapReduce jobs | ||||
Set up schedules (fair scheduler, capacity scheduler, job placement) | ||||
Set up snapshots and mirrors | ||||
Configure NICs | ||||
Set up monitoring tools | ||||
Set up security | ||||
Create a disaster recovery plan |