Parallel full table scan

Hi folks,

I am evaluating the features of cockroach to maybe migrate data from Cassandra to Cockroach.
One thing I was wondering:

Does cockroach support parallel full table scans? This may be pretty useful for large batch processes or map-reduce-like jobs.
E.g. In Cassandra this can be done by splitting the DHT token ring in $x slices and process each of them in parallel without any overlapping.
Is something like that possible in Cockroach?

Thanks for your help!

It depends on what you want to do with the results of this scan. If you can express your processing as a SQL query (such as SELECT category, AVG(duration) FROM tbl GROUP BY category), then CockroachDB’s distributed SQL engine can do this work in a distributed fashion. If not (so you’re not talking about a single table scan but a bunch of separate scans that collectively cover the whole keyspace), CockroachDB doesn’t currently have any tools to facilitate this. For now the best way to do this is to use your own knowledge of your key distribution to construct the appropriate queries. In the future we’ll probably expose some metadata that will let you ask CockroachDB for the appropriate boundary points.

I actually meant to scan the whole keyspace for a single table in a way to
slice the keyspace into n (mostly) equally distributed parts.

Your response answered my question - what I wanted is not supported. It
would still be possible by using an equally distributed and indexed hash
field but it’s a rather dirty workaround.

Thanks anyway!