We have a Tomcat running and to collect JMX logs is it mandatory to have Heavy forwarder installed.?I know we cannot install JMX add-on for UF. Is there any alternative method other than having HF installed.?and catalina,localhost,and manager logs can be collected through UF right.?correct me if I am wrong
from installation guide https://docs.splunk.com/Documentation/AddOns/released/JMX/Distributeddeployment it is said
While it is possible to use the add-on to poll remote JMX management servers, Splunk recommends installing this add-on on a heavy forwarder that is installed on the same host as the central JMX server. This deployment configuration simplifies your configuration because you can call the port locally, you do not need to use certificates, and you can use the Java version that is local to the machine.
based on that a HF is the correct place. I suppose that there are some python parts which needs HF instead of UF. Of course there are some ways to run python also on UF, but much better way is to use HF.
It’s up to you. It seems that some tasks are easier if those are local connections instead of remotes, but both are doable. If you have automation to install and manage HFS, then local installations are probably easier to manage in the long run. If not then centralized way could be easier as all configurations are in one place, even the first installation takes more effort.
Thanks for your help. Last query, I am using UF on my client machine where Tomcat is installed and forwarding Tomcat logs to Splunk indexer and search head. How to make use of the Splunk add-on for Tomcat in this case. Do I need JMX URL here.? Should I make any configuration settings in add-on or it will just parse the logs with the defined sourcetype?