LC/0/5/CPU0:Oct 6 10:55:05 BKK-7: gsp[140]: %OS-gsp-3-MSG_ENQUEUE_ERROR : Process pkg/bin/plim_1p_oc768 (pid 81988) not picking up messages, 23791 msgs dropped. 20482 msgs pending, taking 4632008 bytes

[ปลาวาฬทราย RMUTT CPE IX]

I found this log continue occur on CRS-1 Version 3.6.3. How should I do?

LC/0/5/CPU0:Oct  6 10:55:05 BKK-7: gsp[140]: %OS-gsp-3-MSG_ENQUEUE_ERROR : Process pkg/bin/plim_1p_oc768 (pid 81988) not picking up messages, 23791 msgs dropped. 20482 msgs pending, taking 4632008 bytes

Thank you very much.

[xthuijs]

hi tigma, it means that the linecard in slot 5 is not able to process messages from GSP (group services protocol). GSP is a multicast delivery type of methodology for the RP to send messages like FIB updates to multiple nodes via the EOBC (ether out of band).

This is likely a sw bug, but which one precisely is hard to say. This message is just the result (headache) for which the root cause (eg a heavy night out drinking) is not presented unless we have more logs surrounding this.

It may be hard to go back in time when it started since 20k messages have been enq'd already waiting process with a subsequent 24k dropped. It seems to hold quite a bit of memory also.

You could try to process restart plim_1p_oc768 (pid 81988) on LC 5.

or possibly reload the LC

xander

[ปลาวาฬทราย RMUTT CPE IX]

I don't see the process to restart.

RP/0/RP0/CPU0:u1#sho processes | inc pli

JID   TID CPU Stack pri state        TimeInState    HR:MM:SS:MSEC   NAME
399    1    1   40K  10 Receive        0:00:00:0270    0:00:01:0149 vi_config_replicator
399    2    0   40K  10 Receive     1507:30:54:0156    0:00:00:0002 vi_config_replicator
399    3    1   40K  10 Receive        0:23:36:0686    0:00:00:0768 vi_config_replicator
399    4    0   40K  10 Receive      498:07:59:0512    0:00:00:0042 vi_config_replicator
RP/0/RP0/CPU0:u1#

RP/0/RP0/CPU0:u1#sho processes | inc 988
134    6    1   48K  10 Nanosleep      0:00:02:0988    0:00:00:0735 cethha
RP/0/RP0/CPU0:u1#

The log occurs every 5 minutes. Then I try to reload LC, the log disappeared.

Thank you very much.

[xthuijs]

ah yeah the process would be on the linecard, it would have been show proc loc 0/5/cpu0 pipe include.

any case the issue seems resolved with the lc reload. if this happens again we'll need to investigate some more.

cheers

xander