ike: Restart inactivity counter after doing a CHILD_SA rekey

When doing a rekey for a CHILD_SA, the use counters get reset. An inactivity
job is queued for a time unrelated to the rekey time, so it might happen
that the inactivity job gets executed just after rekeying. If this happens,
inactivity is detected even if we had traffic on the rekeyed CHILD_SA just
before rekeying.

This change implies that inactivity checks can't handle inactivity timeouts
for rekeyed CHILD_SAs, and therefore requires that inactivity timeout is shorter
than the rekey time to have any effect.
This commit is contained in:
Martin Willi 2013-11-07 09:21:02 +01:00
parent 763e035335
commit d048a319df
2 changed files with 6 additions and 3 deletions

View File

@ -386,7 +386,9 @@ retransmission timeout applies, as every exchange is used to detect dead peers.
.TP
.BR inactivity " = <time>"
defines the timeout interval, after which a CHILD_SA is closed if it did
not send or receive any traffic.
not send or receive any traffic. The inactivity counter is reset during CHILD_SA
rekeying. This means that the inactivity timeout must be smaller than the
rekeying interval to have any effect.
.TP
.BR eap_identity " = <id>"
defines the identity the client uses to reply to an EAP Identity request.

View File

@ -73,12 +73,13 @@ METHOD(job_t, execute, job_requeue_t,
{
if (child_sa->get_reqid(child_sa) == this->reqid)
{
time_t in, out, diff;
time_t in, out, install, diff;
child_sa->get_usestats(child_sa, TRUE, &in, NULL, NULL);
child_sa->get_usestats(child_sa, FALSE, &out, NULL, NULL);
install = child_sa->get_installtime(child_sa);
diff = time_monotonic(NULL) - max(in, out);
diff = time_monotonic(NULL) - max(max(in, out), install);
if (diff >= this->timeout)
{