Hi all,
In a very large environment (>6000 agents) the client is considering the option of using SQL AlwaysOn for replication of the OperationsManager DB to a secondary Data Center for Disaster Recovery. However, due to the envisaged high load and storage requirements for the Data Warehouse, they're considering excluding this from the replication and in the event of DR just restoring it to the secondary DC from backup. The thinking being, in a DR event it's the OpsMgr DB that is required to be brought online in a timely fashion (i.e RTO <1hr) to ensure visibility of enterprise health, however the DataWarehouse has a much lower RTO, and can be restored from backup within 24hrs. This will reduce considerable network replication between the DC's for this highly utilised DB.
- Would this be a supportable recovery option, or must the DW be restored with the OpsMgrDB?
- If it can be considered, is there a hard limit of how much data the Management Servers can store locally before things start to break, or is it purely driven from available disk space on the MS's?
- Apart from the obvious continual errors from the MS's complaining about the DW being offline, what potential other functions may break until the DW is brought online (DA's, SLA Dashboards etc?)
Cheers,
E