This gem is intended to replace the existing and unmaintained gelf-rb https://github.com/graylog-labs/gelf-rb - since the project went silent for a few years and there seems to be no intention in continuing any kind of support (graylog-labs/gelf-rb#93). So we decided to not just fork but set up a new gem. Since we needed support for http(s) transports as well we are mirroring this fork (https://github.com/christianrolle/gelf-rb) instead of the original project because we wanted to keep all of the projects and maintainers history.

• 3.1.1 is the latest original gelf-rb version ()
• 3.2.1 is the first http(s) supporting version from here https://github.com/christianrolle/gelf-rb
• 4.0.0 we do start from here with our versioning, improvements and updates

This allows you to send arbitary messages via UDP to Graylog.

n = GELF::Notifier.new("localhost", 12201)

# Send with custom attributes and an additional parameter "foo"
n.notify!(:short_message => "foo", :full_message => "something here\n\nbacktrace?!", :_foo => "bar")

# Pass any object that responds to .to_hash
n.notify!(Exception.new)

The recommended default is to send via UDP but you can choose to send via TCP like this:

n = GELF::Notifier.new("127.0.0.1", 12201, "LAN", { :protocol => GELF::Protocol::TCP })

Note that the LAN or WAN option is ignored for TCP because no chunking happens. (Read below for more information.)

The Gelf::Logger is compatible with the standard Ruby Logger interface and can be used interchangeably. Under the hood it uses Gelf::Notifier to send log messages via UDP to Graylog.

logger = GELF::Logger.new("localhost", 12201, "WAN", { :facility => "appname" })

logger.debug "foobar"
logger.info "foobar"
logger.warn "foobar"
logger.error "foobar"
logger.fatal "foobar"

logger << "foobar"

Then WAN or LAN option influences the UDP chunk size depending on if you send in your own network (LAN) or on a longer route (i.e. through the internet) and should be set accordingly.

Since it's compatible with the Logger interface, you can also use it in your Rails application:

# config/environments/production.rb
config.logger = GELF::Logger.new("localhost", 12201, "WAN", { :facility => "appname" })

• Fork the project.
• Make your feature addition or bug fix.
• Add tests for it. This is important so nothing breaks in a future version unintentionally.

• Feel free to open up pull request

• install docker with compose

You can either run tests against the ruby versions defined in docker-compose.yml:

docker compose up

Or you can run a specific test:

docker compose up ruby26

• The VERSION file has to be updated to the new version number because it is used in gelf_redux.gemspec to have a generic way of bumping the version
• CHANGELOG should be updated with a few hints about the release
• a release tag should be created from main branch after all PRs that will be part of the release have been merged
• creating a tag for the version is mandatory to be able to come back later for a fix-release if needed, example:

git tag v4.2.0
git push origin main --tags

There is no automated way of pushing to rubygems, so we have to run these commands on a developers machine:

gem build *.gemspec
gem push *.gem

There is a github action for automated releases which will run as soon as a version tag has been pushed