Dec 11 19:17:40 XXXX chassism[1214]: JTASK_SCHED_SLIP_KEVENT: 21 sec 494516 usec kevent block
Dec 11 19:17:40 XXXX sfid[1215]: JTASK_SCHED_SLIP_KEVENT: 21 sec 495211 usec kevent block
Dec 11 19:17:40 XXXX eswd[1225]: JTASK_SCHED_SLIP_KEVENT: 21 sec 571357 usec kevent block
Sched slip can happen if a process was hogging the CPU for certain interval, in our case 21 sec.
So during that specific time the CPU cycles were not free to process the packets from the sfid, chassism, eswd, rpd. ntpd numerous demons.
The Lacp is a kernel traffic and should flow through the SFID towards the kernel, and as this specific process encountered a slip, it was not able to pass the lacp pdus towards the kernel, resulting in lacp session drop.
The lacp by default is perodic fast means 1 pdu /sec, so if we configfure slow it would be 10/ sec. as the slip is for 21 sec, this could go unnoticed as miss of 3 PDUs could be session down.
As it is a high CPU related issue, we would suggest to use an external monitoring tool to monitor the cpu usage.
Thanks,
Ramesh.G