second part.
Now backup is completly frozer,
here top:
last pid: 737; load averages: 0.97, 0.43, 0.17up 0+00:02:51
11:30:33
25 processes: 5 running, 20 sleeping
CPU states: 0.3% user, 0.3% nice, 10.2% system, 77.2% interrupt, 11.9%
idle
Mem: 13M Active, 7404K Inact, 10M Wired,
Hmmm, that's a seriously high interrupt load. How much traffic
goes through this box? What type of NICs and CPU do the boxes
have?
--BillOn 8/31/05, Rodolfo Vardelli [EMAIL PROTECTED] wrote:
second part.Now backup is completly frozer,here top:last
pid: 737;load
averages:0.97,0.43,0.17up
On 8/31/05, Rodolfo Vardelli [EMAIL PROTECTED] wrote:
Bill Marquette ha scritto:
standard soekris 4801 + 2 lan
traffic? in that moment zero
and it happens only with carp+load balance with two boards
When you say load balancing do you mean arp balancing?
Scott
There was a bug in previous versions that would send the machine into
a interrupt storm due to a route bug. Your on a later version than
this so I really dont know what to say here.
Scott
On 8/31/05, Rodolfo Vardelli [EMAIL PROTECTED] wrote:
Scott Ullrich ha scritto:
sorry, I meant
Scott Ullrich ha scritto:
In my opinion the bug is still here.
The light (that indicate traffics) of the NIC (wan1 and wan2), are still.
And after I halt the secondary, and the webgui of primary is usuable,
the traffic of the interfaces (1 and 2) is more than 100 Megabytes
regards
Rodolfo
Rodolfo Vardelli ha scritto:
not really solved, it seems that every broadcast pass to other interface:
18:35:30.405454 IP 192.168.9.227.netbios-ns 192.168.9.255.netbios-ns:
NBT UDP PACKET(137): QUERY; REQUEST; BROADCAST
18:35:30.405726 IP 192.168.9.227.netbios-ns 192.168.9.255.netbios-ns:
Perhaps block this rule on the LAN interface before the ALL all?
On 8/31/05, Rodolfo Vardelli [EMAIL PROTECTED] wrote:
Scott Ullrich ha scritto:
What was generating the traffic?
it was a normal broadcast to port 137-138 in network 9.0 that for some
strange reason pass to other network.