Thoughts on sql tagging and targetting

We are just getting started with octopus and we’re trying to figure out the bets way to do out tagging for sql in order to target the correct Db servers. We have a mixture of sql servers that are running standalone, clustering, and mirroring technology. How are people doing their deployments with situations like these? Are you actually installing tentacles on the sql machines or are you running the sql scripts from a remote machine? It seems like the latter may work better for us since we can just use the host name we have configured for the database rather than worrying about which server is hosting which database. Just wondering if anyone could share their strategies to get some ideas. Thanks


The most common approach is to make the database connection string used for deployments into a (sensitive) project-level variable that is scoped to each environment.

That way, just one Tentacle can be used.

If you need to use different Windows credentials for each environment, a role-tagged Tentacle per environment works best (i.e. install the Tentacle next to the SQL Servers).

Hope this helps,