I am creating a parser/converter from the Linux Audit format. As I was studying the format, looking at examples and reading the documentation I stumbled upon a problem.
Can I be sure that the field names inside a single record are unique?
For example, is a record like this one is legal / appear in real world implementations:
type=TYPE msg=audit(1.002:3): msg="the first msg field" msg="the second msg field"
The second related question is whether I can there are there will be only one pid
in an event? For example, is this event is legal / appear in real world implementations:
type=TYPE1 msg=audit(1.002:3): pid=0 msg="texthere" type=TYPE2 msg=audit(1.002:3): pid=0 msg="differenttexthere"
auditd
as the userspace daemon mostly. Linux Audit is a much wider tag which might include questions about the standard itself. I am working a lot with the Linux Audit format documentation lately and I've been missing a proper tag for my questions.audit
seems to be too general andauditd
is not really the tag I need. This is why I decided to create thelinux-audit
tag.auditd
,auditctl
and other related tools). Is there a difference between the Linux audit framework and the Linux audit standard? Please chime in on the meta thread.