October 15, 2015 at 12:08 pm
All of what I ask about below is based upon enterprise edition
I currently have 1 2008r2 active/passive cluster. I have another 2008r2 server that is log shipped to (one primary database) and also houses another instance with SSAS and reporting server. 256 gig of ram on this one. We are getting a replacement active/passive cluster that I was going to move/migrate to 2014. I had wanted to add another node but our new MS software agreement adding another server to be part of AG would cost 250K just for the license (an no, VM is not an option in this solution)
SO I have my walmart solution, is it feasible.
Install new active/passive cluster with 2014, migrate current db to that box.
Upgrade current sharepoint/reporting server/SSAS (also has log shipping copy) to 2014 and remove log shipping.
Add another instance to this same server and add as AG read only version of production database from
cluster and use for reporting
This way I dance around the extra license I need. This server with 2 instances does have 256 gig or RAM, and is pretty underutilized. Feasible?
October 16, 2015 at 4:07 am
tcronin 95651 (10/15/2015)
Install new active/passive cluster with 2014, migrate current db to that box.
This will have a Failover Cluster Instance of SQL Server, correct?
tcronin 95651 (10/15/2015)
Upgrade current sharepoint/reporting server/SSAS (also has log shipping copy) to 2014 and remove log shipping. Add another instance to this same server and add as AG read only version of production database fromcluster and use for reporting
What edition of sql server are you installing here?
You're mentioning Availability groups and this will require Enterprise edition.
tcronin 95651 (10/15/2015)
This way I dance around the extra license I need. This server with 2 instances does have 256 gig or RAM, and is pretty underutilized. Feasible?
The server still needs to be licenced, it will have a production instance of sql for sharepoint running and you also want to create another instance to add as a readable replica in an AG
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
October 16, 2015 at 4:07 am
Obviously you need to contact your licensing vendor for the definitive statement on your situation, but do be aware that if you plan to run any kind of DBCC command on the secondary or any backups on the secondary, that is considered querying, and requires a license.
October 16, 2015 at 4:08 am
Beatrix Kiddo (10/16/2015)
Obviously you need to contact your licensing vendor for the definitive statement on your situation, but do be aware that if you plan to run any kind of DBCC command on the secondary or any backups on the secondary, that is considered querying, and requires a license.
The fact it will be a readable secondary overrides all this
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
October 16, 2015 at 4:10 am
Yep- I was just covering all bases because the original post isn't all that clear.
October 16, 2015 at 6:37 am
to answer them all
1 active/passive cluster (enterprise edition) - 2 nodes one shared storage
another physical server which would have 2 instances, one with SSRS and SSAS, the other instance would be
basically a read-only AG (replace current log shipping). This is physical box enterprise edition. All servers are server proc based. It does seem after getting some feedback from other people that this may be a complicated mousetrap and that the AG is really at the server level not the instance level
October 16, 2015 at 8:53 am
the AG is really at the server level not the instance level
What do you mean by that? AOAGs rely on an underlying Windows Server Failover Cluster, but that doesn't mean that AGs are more of a server thing than an instance thing.
October 18, 2015 at 10:59 am
tcronin 95651 (10/16/2015)
after getting some feedback from other people that this may be a complicated mousetrap and that the AG is really at the server level not the instance level
Feedback from who?
An availability group that contains an FCI and a standalone instance is perfectly acceptable. I would suggest you read my stairway to ha on this site.
The third node will need to be joined to the existing cluster, leave all shared storage zoned off from this third node.
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
Viewing 8 posts - 1 through 7 (of 7 total)
You must be logged in to reply to this topic. Login to reply