|
|
|
|
|
|
by xchaotic
1649 days ago
|
|
|
I work with a 'nosql' database vendor and I take care of customers with PBs of data doing this.
A very common pattern is having a DR setup - i.e. two separate clusters, with the database replicated, then breaking that replication, upgrading DR and if it's successful promoting that to acting prod and upgrading production (cluster #1) and re-enabling replication. |
|
|