Cassandra

Analytics tools#

Datastax bulk loader#

Although the datastax bulk loader is built for Cassandra, it can also be used with ScyllaDB. With that installed, you can execute some long running sql that fails to run in cqlsh, eg. ./dsbulk unload -url ~/tus-old -query "select persistence_id, ser_id from tus_snapshots.snapshots_with_ttl where ser_id > 1 allow filtering" starts a job that executes the query and exports results to a csv file.

Spark cassandra connector#

There is a spark cassandra connector, which also works for scylla as well. Using it we are able to execute sql queries in spark against scylla tables. It's very useful for queries on big tables (that may run for days).