Dear,
after upgrade of our 10.2.0 CU11 FO pair to 10.3.0 CU3 we experienced a failing Cassandra DB Size BPA on both agents. The "localhost" was unable to resolve according the BPA script:
We tried pinging the "localhost" on the agent which was working fine. Also retrieving historical data worked fine which indicated the correct connection to the cassandra database.
In order to fix the BPA we changed the <DBServer> value in the db.xml from "localhost" to "127.0.0.1" and it helped.
Why did this happen? Should we simply leave the "127.0.0.1" in the db.xml?
Seems like a bug inside of the BPA as it is trying to do a nslookup for localhost to resolve it to an IP address. My advice would be to report this as a software issue with our Techsupport (techsupport@skyline.be).