|Persisting Table Data to GemFire XD Disk Stores / Recovering Persistent Data at Startup|
When you start a GemFire XD cluster with disk stores, it is recommended that you start all members that have persisted data at roughly the same time. Also enable network partition detection to avoid persisting inconsistent data during segmentation.
When members with persistent data boot, they determine which have the most recent table data, and they initialize their caches using that data. If you do not start persistent data stores in parallel, some members may hang while waiting for other members to start.
#!/bin/bash # Start all local GemFire XD members to waiting state, regardless of which member holds the most recent # disk store files: ssh servera "gfxd locator start -dir=/locator1 -sync=false -enable-network-partition-detection=true" ssh serverb "gfxd server start -client-port=1528 -locators=localhost -dir=/server1 -sync=false -enable-network-partition-detection=true" ssh serverc "gfxd server start -client-port=1529 -locators=localhost -dir=/server2 -sync=false -enable-network-partition-detection=true" # Wait until all members have finished synchronizing and starting: ssh servera "gfxd locator wait -dir=/locator1" ssh serverb "gfxd server wait -dir=/server1" ssh serverc "gfxd server wait -dir=/server2" # Continue any additional tasks that require access to the GemFire XD members... [...]
See Handling Missing Disk Stores for more information.