November 26, 2014 at 11:29 am
Hi,
We are planning to 2014 migration in few days.
I need help in configuration.
ServerA----- ServerA1
ServerB---- ServerB1
In serverA we have 5databases. And making 5databases as availability group. The replica is ServerA1
In server B we have 3 databases. And the making those 3 databases as an availibility group. The secondary replica is ServerB1.
What is the best option to configure the quoram drive in this situation.
Also Server A1 & Server B1 also we use for reporting purposes.
We have some sensitive data. Is it possible to delete the data while reading the data?
How the memory optimization feature work with always on?
November 26, 2014 at 1:33 pm
ramana3327 (11/26/2014)
Hi,We are planning to 2014 migration in few days.
I need help in configuration.
ServerA----- ServerA1
ServerB---- ServerB1
In serverA we have 5databases. And making 5databases as availability group. The replica is ServerA1
In server B we have 3 databases. And the making those 3 databases as an availibility group. The secondary replica is ServerB1.
What is the best option to configure the quoram drive in this situation.
Can you provide more detail around the Windows Server Failover Cluster configuration, are all nodes members of the same cluster?
Are the nodes separated geographically?
ramana3327 (11/26/2014)
We have some sensitive data. Is it possible to delete the data while reading the data?
Readonly is just that, so all data manipulation will occur at the Primary.
ramana3327 (11/26/2014)
How the memory optimization feature work with always on?
Please read the following MS article, it talks at length around readable secondaries
http://msdn.microsoft.com/en-us/library/ff878253.aspx
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
November 27, 2014 at 10:08 am
Be sure you test the in-memory stuff on a non-production server. Once you create a database as in-memory, it has to stay that way forever. You don't want to do this to a production database until you're absolutely certain that you need it.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
November 30, 2014 at 1:39 pm
Hi Perry thanks for your link.
The client have some business agreement. ServerA databases never mix with ServerB database. Also we are worrying about the load also. If we keep all databases in one server, it may not handle the load. We should have to face performance issues
So surely, we need to have 2 windows failver cluster.
ServerA & Server A1 is together one windows failover cluster, ServerB & ServerB1 is another windows failover cluster.
December 1, 2014 at 6:45 am
ramana3327 (11/30/2014)
Hi Perry thanks for your link.The client have some business agreement. ServerA databases never mix with ServerB database. Also we are worrying about the load also. If we keep all databases in one server, it may not handle the load. We should have to face performance issues
So surely, we need to have 2 windows failver cluster.
ServerA & Server A1 is together one windows failover cluster, ServerB & ServerB1 is another windows failover cluster.
Nope, not supported, all servers must be part of the same windows server failover cluster. You can create an alwayson group spanning replicas across different clusteres but its for migration purposes only.
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
December 2, 2014 at 5:22 pm
I am re-framing my question about the sensitive data.
Currently we are using 3 production servers. In one production server we have one database has sensitive data.
Suppose we have Server A, Server B and Server C.
ServerA has one database (TestDB) with sensitive data.
There is one application connecting to Server C. That application needs to read the non-sensitive data from the database TestDB.
So right now we are taking that TestDB full backup everyday and restoring that database in ServerC then doing some updates and delete commands to hide the sensitive data before that application connect to the ServerC.
Now we are planning to migrate to Sql 2014 and setting up always on read only copy.
They don't want to run the back everyday for TestDB in ServerA.
They wants to read the data for ServcerC directly using always on.
December 3, 2014 at 5:48 am
ramana3327 (12/2/2014)
I am re-framing my question about the sensitive data.Currently we are using 3 production servers. In one production server we have one database has sensitive data.
Suppose we have Server A, Server B and Server C.
ServerA has one database (TestDB) with sensitive data.
There is one application connecting to Server C. That application needs to read the non-sensitive data from the database TestDB.
So right now we are taking that TestDB full backup everyday and restoring that database in ServerC then doing some updates and delete commands to hide the sensitive data before that application connect to the ServerC.
Now we are planning to migrate to Sql 2014 and setting up always on read only copy.
They don't want to run the back everyday for TestDB in ServerA.
They wants to read the data for ServcerC directly using always on.
Is the sensitive data just a few columns in some tables?
In this case you could create a series of Views that exclude the sensitive columns. These would need to be created at the Primary and would then be available in the secondarys
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
December 6, 2014 at 7:09 am
Perry, I am not getting clarity how these availability groups works for reporting.
I do have another doubt
We are planning to have only one secondary replica (only one read only copy). That we want to use source to Reporting. For reporting we don't need to mask any sensitive data. But for the application that connecting to the server C we have to avoid sensitive data.
We need to avoid patient data only
In that case, how the creating views on primary server will work.
Do we need to create a job to execute that views and a table to store the views result?
December 6, 2014 at 12:00 pm
When you create the view on the primary its mirrored to any secondarys just as it would in database mirroring and log shipping.
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
Viewing 9 posts - 1 through 8 (of 8 total)
You must be logged in to reply to this topic. Login to reply