Service Mapping and MS SQL databases , instances and cluster

[stephane_P]

Anyone worked on Service Mapping application patterns ? I am making good progress but still have a few challenges, one of them is related to databases:

1. when discovering an MS SQL instance, I can connect using the host and instance name but unfortunately, it automatically discovers all the databases on that MS SQL instance and shows them as inclusion. I could not find a way to only show the specific databases instances used and I am wondering if there is a way to do that ( maybe with an applicative credential to connect to the MS SQL database but then how can I exclude those database already discovered?)
2. Applications often connect to a MS SQL cluster but there is no such endpoint and I am forced to point to a single host + instance pair. It would have been nice to have the pattern work so that, like a load balancer, it recognize the cluster and cluster nodes where this database instance can run. Anyone  faced this and has a suggestion?

[Rob50]

Hi Stephane,

Did you get a solution to this?

Thanks,

R

[AjinA]

Yes this is a common challenge when working with Service Mapping patterns and MS SQL discovery. By default when you provide host + instance credentials, the probe discovers all databases in that instance because it’s designed to show full visibility. Unfortunately there is no out-of-the-box setting to only include a specific database during discovery. What you can do is control scope using application patterns and credentials with limited permissions. For example, creating a SQL login with access only to the target database(s) will restrict what can be queried during discovery, so the other databases won’t return details. Another option is to filter at the pattern level – adjust the discovery pattern to include only the specific DB names you want.

Regarding clusters, Service Mapping does not treat SQL clusters as load balancers. Instead, you need to define the cluster listener or virtual network name as the entry point if one exists. If not, you’ll have to pick a primary node and then adjust the pattern so that it accounts for failover nodes. Some teams create custom identification rules or extend the OOB patterns so that the mapping shows the cluster object and its members correctly.

And as always if the environment shows recurring corruption or inconsistent discovery results, it’s worth checking the health of the SQL system itself. Tools like Stellar Repair for MS SQL are useful in situations where DB corruption is interfering with discovery or connection attempts, since they can repair damaged MDF/NDF files and restore access cleanly.