[LU-18823] LMR1b: Replicate MDT0000 Services to Other MDS

• Type: Technical task
• Resolution: Unresolved
• Priority: Minor
• Assignee: WC Triage
• Reporter: Andreas Dilger

Description

Replicate and/or allow services running on MDT0000 to other MDTs/MDSes:

• FID Location Database (FLDB) replication, controls FID meta-sequence mapping to MDTs and OSTs, changes very rarely
• Quota Master Target (QMT), that controls quota limits which change only infrequently. Quota usage that is changing continuously can be reconstructed from the MDTs and OSTs directly in case of a failover.
• flock DLM locking, in-memory only, possibly just reconstructed like normal MDT0000 failover and lock recovery, but on another MDS
• group DLM locking, in-memory only, possibly just reconstructed like normal MDT0000 failover and lock recovery, but on another MDS

Issue Links

• Clones: LU-17819 LMR1a: Replicate MGS Services to Multiple MDS
• is related to: LU-19915 keep client configurations on MDTs/OSTs in sync with MGS0
• is related to: LU-17820 LMR2a: Replicate ROOT/ Directory
• is related to: LU-16024 Allow permanently removing an MDT from config
• is related to: LU-16722 MGS config log restructuring
• is related to: LU-17502 distribute flock locking across multiple MDS nodes
• is related to: LU-15414 LMR1: FLDB mirroring
• is related to: LU-15419 LMR1b: Separate Quota Master from MDT0000