druid/docs/content/dependencies/cassandra-deep-storage.md

45 lines
1.9 KiB
Markdown
Raw Normal View History

---
layout: doc_page
---
## Introduction
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](https://github.com/Netflix/astyanax/wiki/Chunked-Object-Store) 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.
## Schema
Below are the create statements for each:
```sql
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;
```
## Getting Started
First create the schema above. I use a new keyspace called `druid` for this purpose, which can be created using the
[Cassandra CQL `CREATE KEYSPACE`](http://www.datastax.com/documentation/cql/3.1/cql/cql_reference/create_keyspace_r.html) command.
Then, add the following to your historical and realtime runtime properties files to enable a Cassandra backend.
```properties
druid.extensions.coordinates=["io.druid.extensions:druid-cassandra-storage:<druid version>"]
druid.storage.type=c*
druid.storage.host=localhost:9160
druid.storage.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`.