bunyan-cassandra
Bunyan stream for saving logs to Cassandra
Installation
npm install bunyan-cassandraExample
Basic example:
var CassandraStream = require('bunyan-cassandra');
var logger = bunyan.createLogger({
src: true,
name: 'name',
streams: [{
level: 'debug',
stream: new CassandraStream({
hosts: ['localhost'],
keyspace: 'mykeyspace',
username: 'myuser',
password: 'mypass',
query: 'INSERT INTO log (id, message) VALUES(uuid(), ?)',
args: ['msg']
})
}]
});
// this mesage will be saved to cassandra
logger.debug('some log message');Advanced example:
var CassandraStream = require('bunyan-cassandra');
var logger = bunyan.createLogger({
src: true,
name: 'name',
streams: [{
stream: process.stdout,
level: 'debug'
}, {
type: 'raw',
level: 'debug',
stream: new CassandraStream({
hosts: ['localhost'],
keyspace: 'mykeyspace',
username: 'myuser',
password: 'mypass',
query: 'INSERT INTO log (id, message, ip, date, user, line, file, func) VALUES(uuid(), ?, ?, ?, ?, ?, ?, ?)',
args: ['msg', 'ctx.ip', 'time', 'ctx.session.user.id', 'src.line', 'src.file', 'src.func'],
callback: function (err, res) {
// this callback is optional
// here you can do something with query results
// or catch possible Cassandra errors
},
transform: function (log) {
// add/remove/edit log object
}
})
}]
});
// ctx can be express request object for example
// it is completely up to you, you just need to configure ``query`` and ``args`` option
// to support provided object
logger.debug({
ctx: {
ip: '127.0.0.1',
session: {
user: {
id: 1
}
}
}
},'this is some log message');Configuration
You can pass following options
hosts
Array with cassandra hosts
default ['localhost']
keyspace
Name of keyspace to use
default 'logs'
username
Cassandra username to use
default 'cassandra'
username
Cassandra password to use
default 'cassandra'
query
Query for inserting data to Cassandra. This will be executed for each log.
default 'INSERT INTO logs (id) VALUES (uuid())'
args
Array of object paths. When new log will be added to Cassandra, module will look
for values on provided paths inside args option, and after that values with
CQL query will be passed to Cassandra, and new record will be added.
For example if your bunyan log entry looks like this:
{
msg: 'this is some log message',
ctx: {
ip: '127.0.0.1',
session: {
id: 2
}
},
level: 20
}
Then if you have CQL query like this:
query: 'INSERT INTO logs (id, msg, ip, sessionid, level) VALUES (uuid(), ?, ?, ?, ?)'
you can populate ? with values if value of your args option is something like this:
args: ['msg', 'ctx.ip', 'ctx.session.id', 'level']
default []
callback
After query is executed callback function will be called (if is defined).
You can catch possible Cassandra errors here, or see result of operation.
default undefined
transform
Each time when new log is received, log will be passed as argument to transform function, so you will have opportunity to modify raw log data, for example add or remove some fields from log.
transform: function (row) {
// let we say we want date with specific format
row.date = moment(row.time).format('YYYY-MM-DD');
return row;
}