This module manages the installation and configuration of tlog for active terminal session recording.
By default, the logs will be recorded to journald
with systems running
systemd
and syslog
otherwise.
See REFERENCE.md for full API details.
This module is a component of the System Integrity Management Platform, a compliance-management framework built on Puppet.
If you find any issues, they may be submitted to our bug tracker.
You can simply include the tlog
class to have the software installed.
To enable automatic session recording, include the tlog::rec_session
class.
You MUST then add all users and/or groups that you want to monitor to the
tlog::rec_session::shell_hook_users
Array.
Note: Groups should be prefixed with a percent sign (%
).
When this is enabled, it will automatically hook into login and interactive
shells based on scripts placed into /etc/profile.d
.
---
tlog::rec_session::shell_hook_users:
- 'root'
- '%administrators'
NOTE: If you want to be 100% certain that all sessions are logged, you should
not rely on this hook but should, instead, set /usr/bin/tlog-rec-session
as
the user's primary shell. This is not feasible in many situations so these
hooks have been provided for the 90% case.
The tlog
project is still evolving so there may be breaking changes that
occur in the future. We highly encourage all users to file feature requests and
bug reports with the upstream project.
iWhen root is set to use the tlog hook, if a user is logged into a system using
a graphical display such as gnome and attempts to su
to root
from more than
one terminal window in the same session, the second su
will not be recorded.
This occurs because, in order to prevent looping, if the session id is the same for both shells, tlog does not start a second recording session. A ticket has been entered for this issue.
The above error does not affect ssh
logins.
To playback tlog from a file, the file must only contain json entries from a single session. The default SIMP implementation of tlog records all sessions with some additional non-json formatted information in a file, causing playback of the raw log file to fail. To generate a usable tlog file for playback, grep and awk can be utilized to filter and format entries for a tlog session. Identify the file containing the raw tlog data. Performing a grep for tlog-rec-session in the logs directory can help locate log files. After identifying the raw log file, examine the contents of the file to identify the rec, a host-unique recording id, for the session to be replayed. The rec can then be used with grep to generate a new file containing only logs from that session in json format:
grep <rec> <raw log file> | awk -F"tlog-rec-session: " '{print $2}' > /tmp/tlog_for_playback
Please read our Contribution Guide.
This module includes Beaker acceptance tests using the SIMP Beaker Helpers. By default the tests use Vagrant with VirtualBox as a back-end; Vagrant and VirtualBox must both be installed to run these tests without modification. To execute the tests run the following:
NOTE: You will need to make sure that the nodesets
can install the tlog
packages from a repository (or install them via beaker
) for the tests to run
successfully.
bundle install
bundle exec rake beaker:suites
Please refer to the SIMP Beaker Helpers documentation for more information.