As you can see from the attached screenshot, my DMA is claiming the database is 0KB in size, I'm assuming that means it's not using the DB, but it should, as long as it's configured, right? 1) can you confirm what the cassandra DB is used for (what does it store, and if it doesn't store it there where does it store it instead, or doesn't it, and the same for the elastic search please, then 2) is there a way to test DB connections in DMA or prove if the connections are OK, and if so why the DB is 0KB?
Cheers, James.
![](https://community.dataminer.services/wp-content/uploads/wp-user-manager-uploads/2024/04/Stokell-James-Profile.jpg)
![](https://community.dataminer.services/wp-content/uploads/wp-user-manager-uploads/2024/01/Edib-Supic-small.jpg)
Hi James,
Sorry for late reply. Can you check SLDBConnection.txt logs?
![](https://community.dataminer.services/wp-content/uploads/wp-user-manager-uploads/2024/04/Stokell-James-Profile.jpg)
I’ve checked and that log file seems to look OK, I see it attempt a connection to and it confirms it was established, and there’s many other lines too, none of which appear to be errors. I had a thought; perhaps it’s working fine, but the DMA Cube isn’t showing the size because one part of DM knows to use the default port (9042) and the monitoring part doesn’t, so can’t get the size info? I’ve updated the DB connection string to specify the 9042 port, I can’t restart the DMA right now to test this, but would make some sense to me. Will update when I’ve done this in an hour.
![](https://community.dataminer.services/wp-content/uploads/wp-user-manager-uploads/2024/04/Stokell-James-Profile.jpg)
I specified the port number and that actually broke the DB connection, so looks to me like some sort of DB bug as to why it’s showing zero data consumption on the DB, could someone take a look at it please?
Hi James,
This seems to be an actual bug for databases other than a local Cassandra (So for database types Cassandra Cluster as wel as Storage as a Service).
The reason is likely that the database structure in these databases is different from the single Cassandra case, as all tables now have their own keyspace rather than everything being in one. DataMiner uses nodetool to fetch that information, but it is likely still using the old request.
I've created an internal task to tackle this issue. This should thus be fixed in one of the future versions.
Does anyone clever have an answer for this please?