I'm using Solaris 11 with the included IPFilter and Squid packages.

After I set up IPNat to do Squid intercept proxy, my system started soft hanging...After a few days of home type use, the system would become unresponsive...I could ping it, and get keyboard echo on the console, but that was about it.

My ipf.conf looks like this:
name="ipf" parent="pseudo" instance=0 fr_statemax=18963 fr_statesize=27091;

I am logging some streams and IPF stats. At the time of the crash, this is what was in the log:


fragment state(in):     kept 0  lost 0  not fragmented 0
fragment state(out):    kept 0  lost 0  not fragmented 0
packet state(in):       kept 401772     lost 0
packet state(out):      kept 95014      lost 32
-----------------------------------
streams allocation:
                                         cumulative  allocation
                      current   maximum       total    failures
streams                  1366      1610     5481543           0
queues                   2261      2682     6543652           0
mblk                    10652     16740    11167294           2
dblk                    10637     18078    94780086          16
linkblk                    13        83          14           0
syncq                      18        50          47           0
qband                       2        62           2           0

4532 Kbytes allocated for streams data


Do the lost packet states and dblk and mblk failures have any relation to my hangs? This system is in a home environment, a busy one, but still, I can't see Solaris having a problem with the small amount of load in a home environment.

Anything I need to tune here?

Thanks...

Reply via email to