NAME¶
recap - dumps periodic information about running applications and resource usage
DESCRIPTION¶
recap is a user-configurable script that can be run once, or run periodically
out of cron or systemd.timers to dump information about running processes and
resource usage. It's useful on servers that have periodic, mysterious
performance anomalies for tracking down what may be going on at the time of
any particular incident.
The values for which reports are generated and how many reports
are stored can be overridden in /etc/recap.conf. The output files
from the script are written to BASEDIR.
- -B, --backup
- This will perform a "backup". It will copy the latest log files
from the last execution to the "backups" directory inside
BASEDIR. The files will be identified by a timestamp in the name.
- -S, --snapshot
- This will perform a "snapshot" report. It will generate a set of
output files in the "snapshot" directory inside BASEDIR outside
of the standard rotation. The files will be identified by a timestamp in
the name.
- -V, --version
- Print version and exit.
FILES¶
/etc/recap.conf
/etc/cron.d/recap
/usr/lib/systemd/system/recap.{service,timer} - Unit files for running recap periodically.
/usr/lib/systemd/system/recap-onboot.{service,timer} - Unit files for running recap on boot.