March 27, 2024 at 11:17 pm
We have two data centers in same city. One is our main site, the other is a disaster recovery site that with minimum set of networks, dns servers are on.
If we want to create SQL server disaster recovery for our 4 servers with 100 databases. I was planning to create a AlwaysOn High availability group for failover, but we can not afford Enterprise licenses for 4 servers, which need to be up at the same time.
In this case if we use Alwayson Failover cluster, not sure if that helps much. I know failover cluster failover at instance level and use shared storage vs Alwayson High availability group uses two sets of data files on two different nodes. So I prefer high availability group for disaster recovery purpose, but since we don't have that budget, if I use Alwayson Failover for instance level failover, will that help on disaster recovery purpose to failover to another site ?(they are on same domain and subnet). The only thing I concerned it is they use same set of data files, if the data files in our main data center, then disaster hits, it won't be available to either site, so it sounds like it is only a good choice for same site disaster recovery, like a disk broken on one hardware.
Other suggestions? thanks
March 28, 2024 at 9:42 am
Create 2 FCI's one in each DC, then link them both with a AG.
You get local automatic failover should a local host fail, and you get manual failover to DR should the whole DC fail.
Keeps you within the standard licensing limitations.
But with 100 databases, means 100 basic AG's means 100 listeners, means 200 IPs, you may want, to look at an alternative like LogShipping.
April 8, 2024 at 11:43 pm
Log shipping is also database based, so 100 db, we need to create 100 jobs on primary node, then copy, restore jobs for each database, that is 2x100 jobs on second node.
still thinking the AlwaysOn availability group may be the best option for us except the cost,
April 18, 2024 at 2:52 pm
Always on availability group with a replica at the DR site is the best choice.
But Depending on your service level agreement (SLA) you can still have :
-Critical databases on Always on Availability Group.
-Non critical databases on standalone instance with full and transaction log backups (every 10 minutes for transaction log backups) moved to the disaster recovery site with regular recovery tests and dbcc checks.
Also, if you are on Server Core, you may be able to go a little bit over the 100 databases limit.This will depend on your CPU\Memory.
April 18, 2024 at 2:53 pm
Always on availability group with a replica at the DR site is the best choice.
But Depending on your service level agreement (SLA) you can still have :
-Critical databases on Always on Availability Group.
-Non critical databases on standalone instance with full and transaction log backups (every 10 minutes for transaction log backups) moved to the disaster recovery site with regular recovery tests and dbcc checks.
Also, if you are on Server Core, you may be able to go a little bit over the 100 databases limit.This will depend on your CPU\Memory.
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply