Obsolete:Monitoring tasks

From Wikitech

Things we need nice, good, visible monitoring for, with nice pretty graphs over time. Many of these are good candidates for Ganglia.

Per-machine:

  • DBs:
    • Replication lag
    • Running processes?
    • Queries/sec
    • Current stats
      • Disk space
      • MySQL
      • MySQL status
      • RAID
      • And normal system stats
    • ganglia-metrics
  • Apaches
    • Hits/sec
    • PHP errors/sec
    • Segfaults/sec
  • Batch processing
    • Job queue processing rate
  • Mail server
    • incoming mails
    • outgoing mails
    • (mails thrown away as junk?)

Per-wiki:

  • Job queue progress
  • Dump progress?