Hi all,
I have a small EDB cluster deployed with a fork of
elephantdb-deploy and I run into the issue of having many versions that, together, are too big for the cluster's memory (disk and RAM). When I remove some older versions there is enough space.
I'm wondering what the current best practises are for cleaning up versions of the shards, on hdfs and on the cluster itself? Is this maybe a configuration thing? E.g. is this built into elephantdb-cascading and/or the server itself?
Cheers,
Jeroen