DB Connect with SHP - multiple Java bridge processes?

[sowings]

If I'm attempting to provide a bit of redundancy / high availability for my database inputs by installing DB connect in a search head pool (SHP), does that mean I will be running multiple Java bridge processes? If one of my SH is unable to execute a scheduled dbmon-tail search, will one of the others in the pool take up the task?

[ziegfried · Accepted solution]

Yes, multiple java bridge processes will run. If you're running that on the same machine, you have to configure different ports for the java bridge on both instances in $SPLUNK_HOME/etc/system/local/java.conf

[bridge]
port = 4712

It's not recommended/supported to run database inputs on DB Connect on a SHP instance. You should use a separate heavy forwarder for this in that case.

[ziegfried]

Correct. There are no integrated means to gain redundancy for database inputs. If this is required then I'd recommend using external means, such as a VMware HA for the heavy forwarder.

[sowings]

A separate heavy forwarder doesn't get me my redundancy, though. The point of trying to run DB connect on the members of a SHP is to have multiple hosts available to do the "dbmon-tail" or "dbquery".

What is the suggested best practice for this, if not to use SHP?