Spotted this kind of error in a local cassandra deployment (no cassandra cluster, DB on the DMA):
what's the impact on the DMA when one or more "storages" go to offload mode?
Is it correct to configure this as a critical alarm? Or should it be some lower severity?
Any steer will be helpful - thanks
Hi Alberto,
These kind of alarms get generated when the DataMiner system loses contact with the database. To prevent data loss we then go into a failsafe mode where data is stored on disk (C:\Skyline DataMiner\Offload\OfflaodedData.sqlite3). Once the conenction is restored the data in these files will be moved to the Database.
As long as these alarms are active it means that the system is either unable to connect to the database or in the process of moving the data to the DataBase.
Since this can impact the speed of the system and cause certain actions to fail, it seem sensible to have these marked as critical. Further investigation will need to be done to figure out why the system went into this failsafe mode. Things to check here are.
Is the cassandra service up, running and reachable
What errors does SLDBConnection.txt show
What errors does SLCassandraHealth.txt show
Alberto
Do you not have any offload configuration of the database? Please note that recent versions of the DataMiner support file-based offload mechanism.
Offload database | DataMiner Docs
"From DataMiner 10.2.0/10.1.1 onwards, it is possible to offload data to a file in the local system cache folder instead of to an offload database. You can do so by going to System Center > Database > Offload and selecting File in the drop-down box at the top. Below this, you can then configure the data offloads (see Configuring data offloads) and set a maximum size for the combined offload files."
Below is an example screenshot for File offload configuration from a DMA.
Thanks.
Thank, Srikanth – I believe this doesn’t apply to the above case – the alarms are related to internal Cassandra storages (no cluster either, so in this case it is the local DB)
Not a problem Alberto. Wording looks so similar, so I thought this may be to do with the file offload configuration.
That’s the beauty of Dojo, Skyline colleagues are always keeping an eye and making sure that we provide right answers to DataMiner users. Thanks.
Thanks for the prompt feedback, Brent – much appreciated
Will follow up with the assigned squad.
It was an odd one as the DB is local, but could see that little by little many storages were impacted – please, is there any section where I can check how many storages are defined for my Cassandra instance?
Is 22 the maximum or would there be more that in this case were not failing?