Wired Intelligent Edge

I get warning messages like
W 10/07/08 08:11:38 system: Out of pkt buffers; miss count: 216683 on my Procurve 2810-48G's quite often.

I'v noticed that it happens most of the times when I reboot a machine. This machine can be connected to the same switch, but can also be connected to another switch on the same network.

When I look at the output of the "show tech buffers" command, I can see that most of the time there are old packets in the buffers. I can imagine that if this ammount is quite high, "out of pkt" can occur. I'm I correct?
An old packet entry looks like:

Current system time: 10/10/08 11:10:34
Buffer addr OwnerName OwnerID When Allocated Lives Operation Alarm
----------- ---------- -------- ----------------- ----- --------- -----
0xa276be90 bcmRX 81f51cf0 10/10/08 09:30:38 0 00000000 F
So the current time is 11:10, but there are still packets from 9:30.

I use 5 VLAN's and those are carried on all the trunks. I have 8 2810-48G switches, all interconnected to 2 2810-24G switches.

I've searched this forum and found some answers about disabling spanning tree, increasing mac-age timeout and enabling/changing Qos, but I guess the real cause is old packets in the buffer.

Can anyone shine a light on this?

Thanks in advance.

 

 

P.S. This thread has been moved from Switches, Hubs, Modems (Legacy ITRC forum) to ProCurve / ProVision-Based. - Hp Forum Moderator

please send me sh tech command print

Here you go

hi


W 10/04/08 11:11:06 system: Out of pkt buffers; miss count: 207539

The number of misses indicates the number of times that an attempt
to allocate the specified message or packet buffer failed.


as for me your problem on trunk 22
ı see very eror log on trk2 member port

*****************************************
W 10/06/08 11:15:18 FFI: port 37-High collision or drop rate. See help.
W 10/06/08 11:15:18 FFI: port 40-High collision or drop rate. See help.
W 10/06/08 12:33:49 system: Out of pkt buffers; miss count: 214268
W 10/06/08 12:34:51 FFI: port 37-High collision or drop rate. See help.
W 10/06/08 12:34:51 FFI: port 40-High collision or drop rate. See help.
I 10/06/08 12:39:55 mgr: SME SSH from 192.168.1.247 - MANAGER Mode
I 10/06/08 13:06:53 mgr: SME SSH from 192.168.1.247 - MANAGER Mode
W 10/07/08 08:03:04 FFI: port 37-High collision or drop rate. See help.
W 10/07/08 08:03:04 FFI: port 40-High collision or drop rate. See help.
********************************************
and buffer overflow switch

I see strange port information for trk22 (int37-40)
very interesting please you look
********************************************
Status and Counters - Port Counters for port 37

Name : ns20 trunk 1 - 1/2
Link Status : Up
Totals (Since boot or last clear) :
Bytes Rx : 0 Bytes Tx : 2,229,267,178
Unicast Rx : 0 Unicast Tx : 640,877
Bcast/Mcast Rx : 0 Bcast/Mcast Tx : 21,590,960
Errors (Since boot or last clear) :
FCS Rx : 0 Drops Rx : 0
Alignment Rx : 0 Collisions Tx : 0
Runts Rx : 0 Late Colln Tx : 0
Giants Rx : 0 Excessive Colln : 0
Total Rx Errors : 0 Deferred Tx : 0
Rates (5 minute weighted average) :
Total Rx (bps) : 0 Total Tx (bps) : 808224
Unicast Rx (Pkts/sec) : 0 Unicast Tx (Pkts/sec) : 0
B/Mcast Rx (Pkts/sec) : 0 B/Mcast Tx (Pkts/sec) : 0
Utilization Rx : 0 % Utilization Tx : 00.08 %





Status and Counters - Port Counters for port 40

Name : ns20 trunk 1 - 2/2
Link Status : Up
Totals (Since boot or last clear) :
Bytes Rx : 0 Bytes Tx : 95,415,373
Unicast Rx : 0 Unicast Tx : 26
Bcast/Mcast Rx : 0 Bcast/Mcast Tx : 1,310,120
Errors (Since boot or last clear) :
FCS Rx : 0 Drops Rx : 0
Alignment Rx : 0 Collisions Tx : 0
Runts Rx : 0 Late Colln Tx : 0
Giants Rx : 0 Excessive Colln : 0
Total Rx Errors : 0 Deferred Tx : 0
Rates (5 minute weighted average) :
Total Rx (bps) : 0 Total Tx (bps) : 808112
Unicast Rx (Pkts/sec) : 0 Unicast Tx (Pkts/sec) : 0
B/Mcast Rx (Pkts/sec) : 0 B/Mcast Tx (Pkts/sec) : 0
Utilization Rx : 0 % Utilization Tx : 00.08 %






*******************************************

I can see either port only TX information

finally trk2 config out of action on switch

*******************************************
show lacp

LACP

PORT LACP TRUNK PORT LACP LACP
NUMB ENABLED GROUP STATUS PARTNER STATUS
---- ------- ------- ------- ------- -------
37 Active Trk22 Blocked No Failure
40 Active Trk22 Blocked No Failure
43 Active Trk23 Up Yes Success
46 Active Trk23 Up Yes Success
******************************************

solution


1-please your make control all trunk cable connection.
2-reconfig trk2 on switch (fristly remove trk2)
3-check trk2 interface on other device (pc server etc.)


cenk

Hi Cenk,

thanks for your reply. The trunk is not active because the two trunks are an active/passive pair to our file server.

But I do acknowledge that it is strange that only Tx information is available.

In my previous post I attached the configuration of the switch in which the problem occurs most, but the "Out of pkt" message appears on each of my switches.

All other switches do not have a LACP trunk to the fileserver, so I guess that the LACP trunk has nothing to do with the problem.

Do you have any other idea what the cause of the "out of pkt" message could be?

Thanks in advance.

This indicates an oversubscription on the port in question, especially when the traffic that passes through that link is high or when it has devices such as servers connected to that interface. This high load of traffic oversubscribes the ports, which exhausts the input buffers and causes the Rx-No-Pkt-Buff counter and input errors to increase rapidly.

If a packet cannot be completely received because the switch is out of packet buffers, this counter is incremented once for every dropped packet. This counter indicates the internal state of the Switching ASICs on the Supervisor and does not necessarily indicate an error condition.

Forget to add, enable Flowcontrol, as a countermeasure for oversubscription,

Good Luck,

Franklyn

@Franklyn: which port in question? Because it happens on all my switches, not just this one with this LACP trunk.

I will investigate if Flowcontrol is an option.

Thanks for your input!

This problem has nothing to do with oversubscription of flow-control.

It's usually an indication that the IP address of the switch is being hit pretty hard, possibly by a lot of multicast or broadcast traffic on the network.

I assume you're running the latest firmware available for these switches?

Otherwise, get the 'show tech' reports from them along with a quick network map, repost here, or open a call directly with HP.

@Matt:

thanks for your input!

the switches' IP address is in a special management VLAN which is not used very often so I guess the problem you indicate does not occur at our site.

We do have a lot of broadcast and multicast traffic, but on different VLANs.

I'm not running the latest firmware, but I scheduled maintainance to update the software next week.

I will get back to this forum to acknowledge if this has fixed the problem.