Scott Ullrich ha scritto:
here top (master):
last pid: 1405; load averages: 2.93, 1.14, 0.48 up 0+00:06:03
16:40:31
31 processes: 5 running, 26 sleeping
CPU states: 1.0% user, 9.7% nice, 6.9% system, 82.3% interrupt, 0.0%
idle
Mem: 12M Active, 7880K Inact, 11M Wired, 24K Cache, 9152K Buf, 87M Free
Swap:
WARNING: R/W mount of denied. Filesystem is not clean - run fsck
PID USERNAME THR PRI NICE SIZE RES STATE TIME WCPU COMMAND
1396 root 1 95 10 6516K 5284K RUN 0:02 0.29% php
681 root 1 76 0 2328K 1584K RUN 0:03 0.00% top
1397 root 1 96 10 6200K 4760K RUN 0:01 0.00% php
540 root 1 76 0 1292K 868K select 0:01 0.00% syslogd
643 root 1 76 0 2988K 2336K select 0:00 0.00% racoon
675 root 1 8 0 1580K 1224K wait 0:00 0.00% login
298 root 1 -58 0 3656K 1748K bpf 0:00 0.00% tcpdump
1403 root 1 8 0 1620K 1128K wait 0:00 0.00% sh
1402 root 1 8 0 1620K 1128K wait 0:00 0.00% sh
297 _pflogd 1 -58 0 1536K 1180K bpf 0:00 0.00% pflogd
680 root 1 20 0 2616K 2000K pause 0:00 0.00% tcsh
299 root 1 -8 0 1188K 688K piperd 0:00 0.00% logger
554 root 1 8 0 1620K 1120K wait 0:00 0.00% sh
543 root 1 76 0 3480K 1960K select 0:00 0.00% mini_httpd
659 root 1 8 0 1300K 984K nanslp 0:00 0.00% cron
674 root 1 86 0 232K 132K nanslp 0:00 0.00%
check_reload_st
676 root 1 8 0 1624K 1092K wait t 0:00 0.00% sh
677 root 124;65 8eck_rel1632K 1160K wait 0:00 0.00% sh
on master webgui is always loading, on backup webgui is sloooooooooow
the light of the carp port is still (high traffic)
regards
Rodolfo
BTW: I meant to look here.. http://cvstrac.pfsense.com/timeline
On 8/29/05, Scott Ullrich <[EMAIL PROTECTED]> wrote:
This used to be a problem on older versions but was fixed in the
0.80.0 series. Look for the milestone that says its fixed and use a
version after that.
On 8/29/05, Rodolfo Vardelli <[EMAIL PROTECTED]> wrote:
Scott Ullrich ha scritto:
the webgui is unusable (also ssh), I will use serial
I noticed another problem with load balance:
I have two wan and a wireless lan connected to the soekris, with load
balance enabled
when I try to reach a wireless access point, every packet is directed to
one of the ip of the outbound pool,
so I have to add a static route to connect to an access point.
regards
Rodolfo
Run top when it happens from a serial console or from exec.php and
show us the output.
Scott
On 8/29/05, Rodolfo Vardelli <[EMAIL PROTECTED]> wrote:
I installed 80.4 on a couple os soekris 4801.
With carp and and outbound load balance enabled.
Everything works, but after a couple of minutes, cpu usage goes to 100%
on master and bakcup.
When I reboot the backup cpu usage return to normal, but again after 2-3
minutes return to 100%.
I suspect carp+outbound load balance.
regards
