rgbriggs
commented
4 years ago On 2020-03-09 17:46, Paul Moore wrote:
If our attempts at causing lost records fail, bring out the "big guns" and try a bit harder. Because of this ability to hit the system a bit harder, we also adjust the number of iterations in an attempt to speed things up if the system is not dropping records using the easy approach.
I hadn't seen this easy/hard hybrid approach but I think this is a good compromise between the two needs.
This change also requires us to add a small sleep before restarting auditd at the end of the test as systemd was complaining about the restarts and failing to (re)start auditd.
It was found this delay was necessary in downstream testing, so yes to this as well.
Suggested-by: Ondrej Mosnacek <omosnace@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
Acked-by: Richard Guy Briggs rgb@redhat.com
pcmoore
If our attempts at causing lost records fail, bring out the "big guns" and try a bit harder. Because of this ability to hit the system a bit harder, we also adjust the number of iterations in an attempt to speed things up if the system is not dropping records using the easy approach.
This change also requires us to add a small sleep before restarting auditd at the end of the test as systemd was complaining about the restarts and failing to (re)start auditd.
Suggested-by: Ondrej Mosnacek omosnace@redhat.com Signed-off-by: Paul Moore paul@paul-moore.com