Unrecoverable data caused by dn node failure

citusdata / citus

Distributed PostgreSQL as an extension

https://www.citusdata.com

GNU Affero General Public License v3.0

10.57k stars 669 forks source link

Unrecoverable data caused by dn node failure #7330

Closed cpp-0722 closed 11 months ago

cpp-0722 commented 11 months ago

Hello, in Citus, I set citus.shard_replicationfactor to 2 When a dn node crashes, I delete the information of that node in the metadata, add a new node, and then execute citus Rebalance_ Start, why is the data of the node that was previously suspended not in a multi copy state

JelteF commented 11 months ago

Using shard_replication_factor is not meant for HA. For that we recommend patroni. shard_replicaton_factor is only meant for quite advanced usecases where you want to increase read throughput.

onurctirtir commented 11 months ago

As Jelte mentioned, this is not an unexpected behavior as shard_replication_factor doesn't aim to provide an HA solution. Feel free to re-open the issue if you have any more questions, thanks!