Druid can use Cassandra as a deep storage mechanism. Segments and their metadata are stored in Cassandra in two tables: index_storage
and descriptor_storage
. Underneath the hood, the Cassandra integration leverages Astyanax. The index storage table is a Chunked Object repository. It contains compressed segments for distribution to historical nodes. Since segments can be large, the Chunked Object storage allows the integration to multi-thread the write to Cassandra, and spreads the data across all the nodes in a cluster. The descriptor storage table is a normal C* table that stores the segment metadatak.
Below are the create statements for each:
CREATE TABLE index_storage ( key text, chunk text, value blob, PRIMARY KEY (key, chunk)) WITH COMPACT STORAGE; CREATE TABLE descriptor_storage ( key varchar, lastModified timestamp, descriptor varchar, PRIMARY KEY (key) ) WITH COMPACT STORAGE;
First create the schema above. (I use a new keyspace called druid
)
Then, add the following properties to your properties file to enable a Cassandra backend.
druid.storage.cassandra=true druid.storage.cassandra.host=localhost:9160 druid.storage.cassandra.keyspace=druid
Use the druid-development@googlegroups.com
mailing list if you have questions, or feel free to reach out directly: bone@alumni.brown.edu
.