Node.js CQL driver for Apache Cassandra
node-cassandra-client is a Node.js CQL 2 driver for Apache Cassandra 0.8 and later.
CQL is a query language for Apache Cassandra. You use it in much the same way you would use SQL for a relational database.
The Cassandra documentation can help you learn the syntax.
Since 0.14.1, the client supports Apache Cassandra 1.2.x in CQL 2
compatibility mode.
By default Cassandra 1.2.x uses CQL 3 so you need to turn the CQL 2
compatibility mode on by passing 'cql_version': '2.0.0' attribute toConnection / PooledConnection constructor in the options object.
Explicitly setting CQL version on a connection is only supported from
Apache Cassandra 1.1 and above so you should only set it if you are
using version 1.1 or above.
For example:
``javascript`
var conn = new Connection({'host': host, 'port': port, 'keyspace': 'Keyspace1', 'cql_version': '2.0.0'});
var pool = new PooledConnection({'hosts': hosts, 'keyspace': 'Keyspace1', 'cql_version': '2.0.0'});
If you use cqlsh which ships with Cassandra 1.2.x or a newer version ofcqlsh which defaults to CQL 3 you need to pass -2 argument to it, otherwise
the client won't be able to read column family definitions (#67).
For example:
`bash`
cqlsh -2 localhost 9160 < my_file.cql
$ npm install cassandra-client
`bash`
npm run-script test
npm run-script lint
node-cassandra-client is distributed under the Apache license.
lib/bigint.js is borrowed
from the Node.js source (which comes from the V8 source).
sys.describeKeyspace('Keyspace1', function(err, ksDef) {
if (err) {
// this code path is executed if the key space does not exist.
} else {
// assume ksDef contains a full description of the keyspace (uses the thrift structure).
}
});
var Connection = require('cassandra-client').Connection;
var con = new Connection({host:'cassandra-host', port:9160, keyspace:'Keyspace1', user:'user', pass:'password'});
con.connect(function(err, con) {
if (err) {
// Failed to establish connection.
throw err;
}
con.execute('UPDATE Standard1 SET ?=? WHERE key=?', ['cola', 'valuea', 'key0'], function(err) {
if (err) {
// handle error
} else {
// handle success.
}
});
});
The Connection constructor accepts the following properties:
host: cassandra host
port: cassandra port
keyspace: cassandra keyspace
user: [optional] cassandra user
pass: [optional] cassandra password
use_bigints: [optional] boolean. toggles whether or not BigInteger or Number instances are in results.
timeout: [optional] number. Connection timeout. Defaults to 4000ms.
log_time: [optional] boolean. Log execution time for all the queries.
NOTE: You'll only get ordered and meaningful results if you are using an order-preserving partitioner.
Assume the updates have happened previously.
`javascript`
con.execute('SELECT ? FROM Standard1 WHERE key >= ? and key <= ?', ['cola', 'key0', 'key1'], function (err, rows) {
if (err) {
// handle error
} else {
console.log(rows.rowCount());
console.log(rows[0]);
assert.strictEqual(rows[0].colCount(), 1);
assert.ok(rows[0].colHash['cola']);
assert.ok(rows[0].cols[0].name === 'cola');
assert.ok(rows[0].cols[0].value === 'valuea');
}
});
PooledConnection() accepts an objects with these slots:
hosts : String list in host:port format. Port is optional (defaults to 9160).
keyspace : Name of keyspace to use.
user : User for authentication (optional).
pass : Password for authentication (optional).
maxSize : Maximum number of connection to pool (optional).
idleMillis : Idle connection timeout in milliseconds (optional).
use_bigints: boolean indicating whether or not to use BigInteger or Number in numerical results.
timeout: : [optional] number. Connection timeout. Defaults to 4000ms.
log_time : [optional] boolean. Log execution time for all the queries.
Timing is logged to 'node-cassandra-client.driver.timing' route. Defaults to false.
Queries are performed using the execute() method in the same manner as Connection,
(see above). For example:
// Writing
connection_pool.execute('UPDATE Standard1 SET ?=? WHERE KEY=?', ['A', '1', 'K'],
function(err) {
if (err) console.log("failure");
else console.log("success");
}
);
// Reading
connection_pool.execute('SELECT ? FROM Standard1 WHERE KEY=?', ['A', 'K'],
function(err, row) {
if (err) console.log("lookup failed");
else console.log("got result " + row.cols[0].value);
}
);
When you are finished with a PooledConnection instance, call shutdown(callback).
Shutting down the pool prevents further work from being enqueued, and closes all
open connections after pending requests are complete.
// Shutting down a pool
connection_pool.shutdown(function() { console.log("connection pool shutdown"); });
Instances of Connection() and PooledConnection() are EventEmitter's and emit log events:
var Connection = require('cassandra-client').Connection;
var con = new Connection({host:'cassandra-host', port:9160, keyspace:'Keyspace1', user:'user', pass:'password'});
con.on('log', function(level, message, obj) {
console.log('log event: %s -- %j', level, message);
});
The level being passed to the listener can be one of debug, info, warn, error, timing and cql. The message is a string and obj is an object that provides more detailed information.
To regenerate Thrift definition files you need to have thrift compiler
installed. You can find information on how to do that on the
Thrift website.
`bash`
git clone git://github.com/apache/cassandra.git
cd cassandra/interface
thrift --gen js:node cassandra.thrift
The Javascript Number type doesn't match up well with the java longs and integers stored in Cassandra.
Therefore all numbers returned in queries are BigIntegers. This means that you need to be careful when you
do updates. If you're worried about losing precision, specify your numbers as strings and use the BigInteger library.
node-cassandra-client supports Cassandra BytesType, IntegerType, LongTime and TimeUUIDType out of the box.BigInteger
When dealing with numbers, the values you retreive out of rows will all be s (be wary of losing precision
if your numbers are bigger than 2^53--you know, like a timestamp).
BigInteger` supports many operations like add, subtract, multiply, etc., and a few others that may come in handy: shift, square, abs, etc. Check the source if you'd like to know more.
We technically have a UUID type, but have not had the need to flesh it out yet. If you find the need to expose more parts of the UUID (timestamp, node, clock sequence, etc.), or would like to implement some operations, patches are welcome.
* Full BigInteger documentation.