Unexpected log in Backbone Matrix X Series


The logs are like this. Also,

Jul 10 13:41:16 iom1 fabricModuleManager: [13:41:16] QE[0]: status changnge QM_ERROR2 = 0x0000000f: Queue age event(s) occurred on queue 7/0x0007

Jul 10 13:51:17 iom16 fabricModuleManager: [13:51:17] QE[1]: status ch\Jul 10 13:51:17 iom16 fabricModuleManager: [13:51:17] QE[1]: status chanange QM_ERROR2 = 0x0000000f: Queue age event(s) occurred on queue 7/0x0007

Jul 10 13:51:17 iom16 fabricModuleManager: [13:51:17] QE[1]: suppressi\
ng logging of any repeat errors for 600 seconds

Jul 10 13:51:24 iom1 fabricModuleManager: [13:51:24] QE[0]: re-enablin\
g logging of all suppressed errors

Jul 10 13:51:24 iom1 fabricModuleManager: [13:51:24] QE[0]: status cha\
nge QM_ERROR2 = 0x0000000f: Queue age event(s) occurred on queue 15/0x000f

3 replies

Userlevel 5
Hello Erhan,

These dont necessarily mean something is wrong. This could be the result of a related application such as traffic shaping.
Its been some time since working with the X - but as I remember, the messaging is related to an M8 FE module and indicate, in this case, packets at the head of queue 15 and queue 7 have been aged out. Any time a packet is in queue for 2 seconds this age event takes place.

The condition may be transient - or it may be a symptom of some other root cause.

Examples:
-may result from any sustained condition involving rate limiting or soft-path forwarding that might lead to packets stacking in buffer; congestion when using large queue depths. One case may be bcast/multicast buffering due to a variety of egress conditions on a group of ports. (ex: 10hdx port and a 10g port on the same vlan egress list)
- may be a symptom of degraded network service due to a root cause such as loop or a flow control interaction.

- sampled netflow and rmon p-cap tend to make these messages more prevalent.

These tend not to be critical messages (notice level) but if you are interested in further investigation I recommend opening a case with our GTAC support team to include more of your environment's specifics.

One mitigation approach could be to make queue sizes smaller.
See release notes and Matrix-X system documentation for default queue structure and configuration.

best regards,
-mike
Hello Erhan,

The QM_Error message indicates that one of the transmit queue's on an IOM1 and IOM16 was aged because a packet could not be transmitted in 2 seconds. The queue is getting backed up for some reason.



Check to make sure that there are no ports in this system running at half duplex, and none are running at 10 mbps.



Two other reasons for the queue backup could be flow-control, or if they modified the traffic-profiles in a manner that can overrun the queues.

Regards,

Douglas Hunt

Extreme Network Support
Thanks for reply.

I will check something related on your answers.

Reply