ApplicationHA 6.1, ESXi 5.5 and SQL 2012 problems
Hi all,
Just wondering if anyone out there might be able to throw a bit of advice my way. Currently evaluating AppHA on VMWare (AppHA 6.1 and ESXi 5.5 hosts) with the view to buying as on paper it does exactly what we want. I've installed the High Availability Console server which has plugged into vCenter nicely, installed the ApplicationHA agent onto a VM running Windows 2012 DC R2 and SQL 2012, again all going smoothly up to this point. Now I get around to testing the agent out. I'm doing this by stopping the sql server and sql agent services on the VM and waiting for AppHA to kick in and restart both services, this is when I'm getting into a little trouble.
Firstly the services do not shut down as cleanly as when the agent is not configured on the VM, a message pops up saying that windows could not stop the SQL server and that the service did not return an error, helpful I know. This does not happen when the agent is not configured on the VM, the services shut down fine.
Secondly when the agent is configured and the services finally shut down, the AppHA plugin for the VM in vCenter displays a faulted status and that the SQL server service is not running and that the SQL agent service could not started. AppHA just sits there and does nothing, it does not restart the services as it should do.
Now it's probably something I've done during setup but being new to the product I cannot see what, I have followed all the install pdf's and have read them over many times, I just cannot see what I may have done wrong as the setup seems simple enough. Or am I testing the product in the wrong fashion? As I'm only in evaluation mode I have no technical support with Symantec so any help with this would be very much appreciated as we are looking at buying this product and at the moment this is the final piece of the test I need to get working (probably the most important part) before I can turn to the bean counters and ask for some cash.
Many Thanks
Lee
Hi Lee,
What is the name of the customer?
Regarding the issue you are facing i.e. the service does not shutdown properly:
This is because the IMF feature was ON for the SQL Service monitoring agent of AppHA.
IMF feature helps an AppHA monitoring agent to detect the state change of the monitored component very quickly (almost instantly).
In this particular case:
SQL has a feature to restart SQL server (or SQL server agent) in case it stops unexpectedly.
We have made IMF compatible with this feature:
If SQL IMF detects that SQL server has gone down it will read the SQL setting and restart service in case Auto restart is set to true.
And that's was happening here. The AppHA agent tries to restart the agent as soon as it detected that SQL service went down (while the service was still going down via the SCM).
To avoid this situation, you can disable the feature of SQL.
Steps to disable SQL auto restart:
1.Ensure SQL server and SQL Server Agent services are running
2.Launch SQL server management studio and connect to localhost
3.Expand localhost node and right click on “SQL Server Agent” node and select “Properties”
4.Uncheck the check boxes –
a.Auto restart SQL server if it stops unexpectedly
b.Auto restart SQL server Agent if it stops unexpectedly
5.Click OK
[Check the attached screenshot]
For SQL monitoring agent to pick-up/refresh this new setting, do the following:
1.Stop VCS:
hastop -local -force
2.Start VCS:
hastart