pino-multi-stream is a wrapper around the pino logger. The purpose of pino-multi-stream is to provide a stop-gap method for migrating from the Bunyan logger. Whereas pino allows only one destination stream, pino-multi-stream allows multiple destination streams via the same configuration API as Bunyan.
Please see the caveats section for some important information regarding the performance of this module.
For Pino v5+
npm install -s pino-multi-stream
For Pino v4 and below:
npm install -s pino-multi-stream@legacy #v3 pino-multi-stream line
pino-multi-stream does not provide the CLI that pino provides. Therefore, you should not install it globally.
var fs = require('fs')
var pinoms = require('pino-multi-stream')
var streams = [
{stream: fs.createWriteStream('/tmp/info.stream.out')},
{level: 'fatal', stream: fs.createWriteStream('/tmp/fatal.stream.out')}
]
var log = pinoms({streams: streams})
log.info('this will be written to /tmp/info.stream.out')
log.fatal('this will be written to /tmp/fatal.stream.out')
The API for pino-multi-stream is the same as that for pino. Please read pino's documentation for full details. Highlighted here are the specifics for pino-multi-stream:
-
The signature for constructor remains the same,
pino(opts, stream)
, but there are a few conditions under which you may get a real pino instance or one wrapped by pino-multi-stream:-
If the
opts
parameter is a writable stream, then a real pino instance will be returned. -
If the
opts
parameter is an object with a singularstream
property then a real pino instance will be returned. If there is also a pluralstreams
property, the singularstream
property takes precedence. -
If the
opts
parameter is an object with a pluralstreams
property, does not include a singluarstream
property, and is an array, then a pino-multi-stream wrapped instance will be returned. Otherwise,opts.streams
is treated a single stream and a real pino instance will be returned.
-
-
The pino options object accepts a
streams
option, as alluded to in then previous item. This option should be an array of stream objects. A stream object is one with at least astream
property and, optionally, alevel
property. For example:var logger = pinoms({ streams: [ {stream: process.stdout}, // an "info" level destination stream {level: 'error', stream: process.stderr} // an "error" level destination stream ] })
Manually create a single multistream
as used internally by the
wrapper:
var fs = require('fs')
var pino = require('pino')
var multistream = require('pino-multi-stream').multistream
var streams = [
{stream: fs.createWriteStream('/tmp/info.stream.out')},
{level: 'debug', stream: fs.createWriteStream('/tmp/debug.stream.out')},
{level: 'fatal', stream: fs.createWriteStream('/tmp/fatal.stream.out')}
]
var log = pino({
level: 'debug' // this MUST be set at the lowest level of the
// destinations
}, multistream(streams))
log.debug('this will be written to /tmp/debug.stream.out')
log.info('this will be written to /tmp/debug.stream.out and /tmp/info.stream.out')
log.fatal('this will be written to /tmp/debug.stream.out, /tmp/info.stream.out and /tmp/fatal.stream.out')
opts
multistream options object. Available options are:
-
dedupe
: Set this totrue
to send logs only to the stream with the higher level. Default:false
dedupe
flag can be useful for example when using pino-multi-stream to redirecterror
logs toprocess.stderr
and others toprocess.stdout
:var pino = require('pino') var multistream = require('pino-multi-stream').multistream var streams = [ {stream: process.stdout}, {level: 'error', stream: process.stderr}, ] var log = pino({ level: 'debug' // this MUST be set at the lowest level of the // destinations }, multistream(streams, { dedupe: true })) log.debug('this will be written ONLY to process.stdout') log.info('this will be written ONLY to process.stdout') log.error('this will be written ONLY to process.stderr') log.fatal('this will be written ONLY to process.stderr')
You can set the level to all streams by changing the level property. It accepts the same parameters as pino. If the level is changed on a child logger, it does not alter the parent streams level. As this is costly operation, we recommend not changing the level for each child logger that is being created.
The behavior of the get accessor changes if { bunyan: true }
is passed
to pinoms. In that case, it implements the
bunyan.level
function.
Note: after 4.1.0 this pino-multi-stream
function was changed according to that of pino
, and after 4.2.0 its API (names of parameters and their options) are in conformity with API of pino
/pino-pretty
, as it is presented here. Those changes are related to pretty-printing options only.
Manually create an output stream with a prettifier applied.
var fs = require('fs');
var pinoms = require('pino-multi-stream')
var prettyStream = pinoms.prettyStream()
var streams = [
{stream: fs.createWriteStream('my.log') },
{stream: prettyStream }
]
var logger = pinoms(pinoms.multistream(streams))
logger.info("HELLO %s!", "World")
The options object may additionally contain a prettifier
property to define which prettifier module to use. When not present, prettifier
defaults to pino-pretty
⇗ (must be installed as a separate dependency).
The method may be passed an alternative write destination, but defaults to process.stdout
.
Prettifying options (after 4.2.0) are to be set like this:
const prettyStream = pinoms.prettyStream(
{
prettyPrint:
{ colorize: true,
translateTime: "SYS:standard",
ignore: "hostname,pid" // add 'time' to remove timestamp
},
prettifier: require('pino-pretty') // not required, just an example of setting prettifier
// as well it is possible to set destination option
}
);
Stern warning: the performance of this module being dependent on the number
of streams you supply cannot be overstated. This module is being provided so
that you can switch to pino from Bunyan and get some immediate improvement,
but it is not meant to be a long term solution. We strongly suggest that you
use this module for only as long as it will take you to overhaul the way
you handle logging in your application. pino-multi-stream
offers close
to zero overhead if there is only one destination stream.
To illustrate what we mean, here is a benchmark of pino and Bunyan using "multiple" streams to write to a single stream:
benchBunyanOne*10000: 703.071ms
benchPinoMSOne*10000: 287.060ms
Now let's look at the same benchmark but increase the number of destination streams to four:
benchBunyanFour*10000: 2249.955ms
benchPinoMSFour*10000: 1017.886ms
And, finally, with ten destination streams:
benchBunyanTen*10000: 4950.301ms
benchPinoMSTen*10000: 3127.361ms