May 6, 2014 at 2:57 am
I have MasterDataServices installed on a server and the database is in an AlwaysOn Availability Group.
The MDM front end currently is set up incorrectly and is referencing NodeA (primary node) for its database.
I want to fix this but im unsure if i should:
-point the MDM front end to the cluster name, so it wont matter what server the databsae is on?
-create an Availability Group Listner, which would also decoupple the database referenced from a particular server.
(note im using MDM here, but that could be any website referencing a database that is in an AlwaysOn Availability Group).
Thanks,
May 6, 2014 at 7:10 am
winston Smith (5/6/2014)
The always on front end currently is set up incorrectly and is referencing NodeA (primary node) for its database.I want to fix this but im unsure if i should:
What front end?
Fix what?
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
May 6, 2014 at 9:59 am
crap, typo, meant to say the MDM front end. ive updated the original post.
May 6, 2014 at 4:40 pm
winston Smith (5/6/2014)
I have MasterDataServices installed on a server and the database is in an AlwaysOn Availability Group.The MDM front end currently is set up incorrectly and is referencing NodeA (primary node) for its database.
I want to fix this but im unsure if i should:
Pointing the web app at the current Primary instance will produce issues if there is an AlwaysOn group failover.
winston Smith (5/6/2014)
-point the MDM front end to the cluster name, so it wont matter what server the databsae is on?
What cluster name are you referring to?
winston Smith (5/6/2014)
-create an Availability Group Listner, which would also decoupple the database referenced from a particular server.(note im using MDM here, but that could be any website referencing a database that is in an AlwaysOn Availability Group).
Thanks,
A listener should be used to provide centralised access to the AlwaysOn availability group so that the moving of the primary role does not present any issues.
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
May 7, 2014 at 1:14 am
Pointing the web app at the current Primary instance will produce issues if there is an AlwaysOn group failover.
Yep, thats the crux of the post, i want to resolve this 🙂
What cluster name are you referring to?
so there is nodeA & nodeB in the cluster. The cluster name is Clust1. I can open SSMS, type in Clust1 in the connection properties and connect to the cluster. I was wondering if in MDM i should referr to this as the instance name, or if i should create an AG Listener.
A listener should be used to provide centralised access to the AlwaysOn availability group so that the moving of the primary role does not present any issues.
And that answers my last question!
New question, whats the benefit of using the AG Listener over the cluster name? after a node failure and failover, the database should still be accessable by either, right?
May 7, 2014 at 3:34 am
winston Smith (5/7/2014)
Yep, thats the crux of the post, i want to resolve this 🙂
ok, so create a listener for your AlwaysOn group
winston Smith (5/7/2014)
so there is nodeA & nodeB in the cluster. The cluster name is Clust1. I can open SSMS, type in Clust1 in the connection properties and connect to the cluster. I was wondering if in MDM i should referr to this as the instance name, or if i should create an AG Listener.
You're connecting to the Windows Cluster Client Access Point right?
This is a really bad idea, quite apart from the fact that this CAP is meant for Cluster management access only!!
Create a dedicated listener for your AlwaysOn group.
winston Smith (5/7/2014)
And that answers my last question!
Good, so get cracking then 😉
winston Smith (5/7/2014)
New question, whats the benefit of using the AG Listener over the cluster name? after a node failure and failover, the database should still be accessable by either, right?
There is no benefit, do not use the Windows Cluster CAP to access the SQL instances, period!
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply