Bonfire is a command line interface to query Graylog searches via the REST API. It tries to emulate the feeling of using tail on a local file.
Examples:
> bonfire -h logserver -u jdoe -@ "10 minutes ago" "*" ... > bonfire -h logserver -u jdoe -f "source:localhost AND level:2" ...
Bonfire usage:
Usage: bonfire [OPTIONS] [QUERY]
Bonfire - An interactive graylog cli client
Options:
--node TEXT Label of a preconfigured graylog node
-h, --host TEXT Your graylog node's host
-s, --tls Use HTTPS
--port INTEGER Your graylog port (default: 12900)
--endpoint TEXT Your graylog API endpoint e.g /api (default:
/)
-u, --username TEXT Your graylog username
-p, --password TEXT Your graylog password (default: prompt)
-k, --keyring / -nk, --no-keyring
Use keyring to store/retrieve password
-@, --search-from TEXT Query range from
-#, --search-to TEXT Query range to (default: now)
-t, --tail Show the last n lines for the query
(default)
-d, --dump Print the query result as a csv
-f, --follow Poll the logging server for new logs
matching the query (sets search from to now,
limit to None)
-l, --interval INTEGER Polling interval in ms (default: 1000)
-n, --limit INTEGER Limit the number of results (default: 10)
-a, --latency INTEGER Latency of polling queries (default: 2)
-r, --stream TEXT Stream ID of the stream to query (default:
no stream filter)
-e, --field TEXT Fields to include in the query result
-x, --template-option TEXT Template options for the stored query
-s, --sort TEXT Field used for sorting (default: timestamp)
--asc / --desc Sort ascending / descending
--help Show this message and exit.
Bonfire can be configured. It will look for a ~/.bonfire.cfg or a bonfire.cfg (in the current directory). The
configuration file can specify API nodes. If no host is specified a node with the name default will be used. You can
also configure queries which can be referenced by starting your query with a colon:
[node:default] host=1.2.3.4 port=12900 endpoint=/ username=jdoe [node:dev] host=4.3.2.1 port=9000 endpoint=/api username=jdoe password=H4rrH4rrB0bba [query:example] query=facility:*foo* AND source:*bar* from=2015-03-01 15:00:00 limit=100 fields=message,name,facility,source
Now you can run queries via such as:
> bonfire --node=dev :example ... runs the example query on the node dev > bonfire :example ... runs the example query on the default node
Cf. the travis script for running tests. Make sure you disable proxies / set
noproxy if you're using a proxy, to be able to reach the test dummy server
(called dummyserver), e.g.:
> env no_proxy=dummyhost python setup.py test
Once you want to release a new version, do the following:
- bring your git tree in order, cut the release, and tag it with the desired version
- install necessary pip packages: > python3 -m pip install –-user –-upgrade setuptools wheel twine
- bundle your release: > python3 setup.py sdist bdist_wheel
- publish it: > python3 -m twine upload dist/*
Note that if you have several different versions in dist, you might want to specify which one you want to publish.
- v0.0.7: Issues fixes, TLS and Proxy support
- Adds support for proxies
- Adds support for https urls
- Add timestamps to the dump format
- v0.0.6: Documentation fix
- Change of README.rst
- v0.0.5: Clean up
- Removed terminal UI ideas
- Added first tests
- Fixed date and time handling with timezones
- Added python3 compatibility
- v0.0.4: Extended documentation & stream access
- Use the first stream the user has access to if no stream is specified and the user has no global search rights
- v0.0.3: Small fixes
- Use accept header in GET requests.
- Fix bug when querying specific fields
- v0.0.1: Initial release
- Limited feature set.