The streamroller npm package is a file-based logging utility designed to help manage log files by supporting automatic rolling of logs based on size or date. It is particularly useful for applications that generate a lot of log data and need to manage disk space efficiently.

What are streamroller's main functionalities?

Rolling file streams based on size

This feature allows you to create a log file that rolls over when it reaches a certain size. In the example, a new log file is created when the current log file reaches 10 MB. The '3' indicates that a maximum of three backup files are kept.

const StreamRoller = require('streamroller');
const stream = new StreamRoller.RollingFileStream('example.log', 1024 * 1024 * 10, 3);
stream.write('This is a log entry');
stream.end();

Rolling file streams based on date

This feature allows for log files to be rolled over based on date patterns. The 'yyyy-MM-dd' pattern means the log file will roll over daily. The 'daysToKeep' option specifies that logs older than 10 days should be deleted.

const StreamRoller = require('streamroller');
const stream = new StreamRoller.DateRollingFileStream('example.log', 'yyyy-MM-dd', { daysToKeep: 10 });
stream.write('This is a log entry');
stream.end();

Other packages similar to streamroller

streamroller

node.js file streams that roll over when they reach a maximum size, or a date/time.

npm install streamroller

usage

var rollers = require('streamroller');
var stream = new rollers.RollingFileStream('myfile', 1024, 3);
stream.write("stuff");
stream.end();

The streams behave the same as standard node.js streams, except that when certain conditions are met they will rename the current file to a backup and start writing to a new file.

new RollingFileStream(filename [, maxSize, numBackups, options])

filename <string>
maxSize <integer> - defaults to 0 - the size in bytes to trigger a rollover. If not specified or 0, then no log rolling will happen.
numBackups <integer> - defaults to 1 - the number of old files to keep (excluding the hot file)
options <Object>
- encoding <string> - defaults to 'utf8'
- mode <integer> - defaults to 0o600 (see node.js file modes)
- flags <string> - defaults to 'a' (see node.js file flags)
- compress <boolean> - defaults to false - compress the backup files using gzip (backup files will have .gz extension)
- keepFileExt <boolean> - defaults to false - preserve the file extension when rotating log files (file.log becomes file.1.log instead of file.log.1).
- fileNameSep <string> - defaults to '.' - the filename separator when rolling. e.g.: abc.log.1 or abc.1.log (keepFileExt)

This returns a WritableStream. When the current file being written to (given by filename) gets up to or larger than maxSize, then the current file will be renamed to filename.1 and a new file will start being written to. Up to numBackups of old files are maintained, so if numBackups is 3 then there will be 4 files:

     filename
     filename.1
     filename.2
     filename.3

When filename size >= maxSize then:

     filename -> filename.1
     filename.1 -> filename.2
     filename.2 -> filename.3
     filename.3 gets overwritten
     filename is a new file

new DateRollingFileStream(filename [, pattern, options])

filename <string>
pattern <string> - defaults to yyyy-MM-dd - the date pattern to trigger rolling (see below)
options <Object>
- encoding <string> - defaults to 'utf8'
- mode <integer> - defaults to 0o600 (see node.js file modes)
- flags <string> - defaults to 'a' (see node.js file flags)
- compress <boolean> - defaults to false - compress the backup files using gzip (backup files will have .gz extension)
- keepFileExt <boolean> - defaults to false - preserve the file extension when rotating log files (file.log becomes file.2017-05-30.log instead of file.log.2017-05-30).
- fileNameSep <string> - defaults to '.' - the filename separator when rolling. e.g.: abc.log.2013-08-30 or abc.2013-08-30.log (keepFileExt)
- alwaysIncludePattern <boolean> - defaults to false - extend the initial file with the pattern
- ~~daysToKeep~~ numBackups <integer> - defaults to 1 - the number of old files that matches the pattern to keep (excluding the hot file)
- maxSize <integer> - defaults to 0 - the size in bytes to trigger a rollover. If not specified or 0, then no log rolling will happen.

This returns a WritableStream. When the current time, formatted as pattern, changes then the current file will be renamed to filename.formattedDate where formattedDate is the result of processing the date through the pattern, and a new file will begin to be written. Streamroller uses date-format to format dates, and the pattern should use the date-format format. e.g. with a pattern of "yyyy-MM-dd", and assuming today is August 29, 2013 then writing to the stream today will just write to filename. At midnight (or more precisely, at the next file write after midnight), filename will be renamed to filename.2013-08-29 and a new filename will be created. If options.alwaysIncludePattern is true, then the initial file will be filename.2013-08-29 and no renaming will occur at midnight, but a new file will be written to with the name filename.2013-08-30. If maxSize is populated, when the current file being written to (given by filename) gets up to or larger than maxSize, then the current file will be renamed to filename.pattern.1 and a new file will start being written to. Up to numBackups of old files are maintained, so if numBackups is 3 then there will be 4 files:

     filename
     filename.20220131.1
     filename.20220131.2
     filename.20220131.3

Keywords

FAQs

What is streamroller?

Is streamroller popular?

Is streamroller well maintained?

Package last updated on 15 Feb 2023

Did you know?

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Install