inbound/outbound packets are few

[shone]

We have two J7000 workstations equipped with same hardware, running HP-UX 11.
Recently I found one of both is slow down, but the led signal is normal and no hardware error messages during boot.
The netstat shows that the packet amount is very few compared to another one.

problem one(j1):
#netstat 1
(lan1)-> input output (Total)-> input output
packets packets packets packets
56542 72326 412759 339525
56547 72331 412771 339536
56548 72332 412776 339541
# netstat -i
Name Mtu Network Address Ipkts Opkts
lan1 1500 192.168.1.0 j1-g 56637 72381
lan0 1500 172.21.0.0 j1 354819 265623
lo0 4136 loopback localhost 2181 2181

normal one(j2):
netstat 1
(lan2)-> input output (Total)-> input output
packets packets packets packets
177924005 116423233 397692663 187251270
177924005 116423233 397692670 187251277
177924005 116423233 397692676 187251283
# netstat -i
Name Mtu Network Address Ipkts Opkts
lan2 1500 192.168.1.0 j2-g 177924048 116423265
lan0 1500 172.21.0.0 j2 219627362 70686435
lo0 4136 loopback localhost 142775 142775

host with '-g' is for 1000Base-SX card.

What is the possible reason for this problem?

Thanks

[Devender Khatana]

Hi,

This does not point to any problems. The utilization also is a factor which will effect this. Also the uptime for both machines can be different causing this differences. You need to collect some more information regarding both systems for us to diagnose this.

How you found it is slow ? What are the applications on the both ? Any recent changes? Any errors is syslog & dmesg etc.

HTH,
Devender

[rmueller58]

Shone,

I'd recommended checking that the switch and the HP NIC parameter values are not in a mismatched state.

If for example your switch is defined as autosense and your HP is defined as 100full, it could be the mismatch is jagging things up a bit.

I don't like autosense.

I'd recommend setting the values in SAM and on the switch port. So you don't have any mismatch.

your problem is actually indicative of that.. I've seen it a lot..