[rbak-nsp] SEOS upgrade from 6.1.3.6 to 6.3.1.1 - problems

[Marcin Kuczera]

hello,

after upgrading SE100 fro 6.1.3.6 to 6.3.1.1 (last 6.3. without BGP 
aggregator problem) we had something like that during system startup:

Oct 17 03:15:42: %DHELPER-3-FILE_ERR: Save file device is no longer
available.
Oct 17 03:15:42: %DHELPER-3-FILE_ERR: Save file unavailable
Oct 17 03:15:42: %IPC-3-ERR: ND-deamon(Thread 7f000205:e03e0006):
ipcSendCommon(): EP 0:0 is dead
Oct 17 03:15:42: %ND-3-GEN_ERR: mbe ism registration failed due to ipc
error -15
Oct 17 03:15:42: %ND-6-INFO: nd mbe registration with ism failed
Oct 17 03:15:42: %IPC-3-ERR: ND-deamon(Thread 7f000205:e03e0006):
ipcSendCommon(): EP 0:0 is dead
Oct 17 03:15:42: %ND-3-GEN_ERR: mbe ism registration failed due to ipc
error -15
Oct 17 03:15:42: %ND-6-INFO: nd mbe registration with ism failed
Oct 17 03:15:42: %IPC-3-ERR: ND-deamon(Thread 7f000205:e03e0006):
ipcSendCommon(): EP 0:0 is dead
Oct 17 03:15:42: %ND-3-GEN_ERR: mbe ism registration failed due to ipc
error -15
Oct 17 03:15:42: %ND-6-INFO: nd mbe registration with ism failed


and when router was up and running:
Oct 17 04:52:03: %BGP-3-ERR: shutdown fd 63 failed: Invalid argument
Oct 17 04:52:12: %BGP-3-ERR: shutdown fd 28 failed: Invalid argument
Oct 17 04:52:23: %BGP-3-ERR: shutdown fd 63 failed: Invalid argument
Oct 17 04:52:23: %BGP-3-ERR: shutdown fd 64 failed: Invalid argument
Oct 17 04:52:36: %HR-3-ERR: shutdown fd 13 failed -- Invalid argument
Oct 17 04:52:43: %BGP-3-ERR: shutdown fd 63 failed: Invalid argument
Oct 17 04:53:02: %BGP-3-ERR: shutdown fd 63 failed: Invalid argument

restart of BGP/HR proces didn't help.


So, you have any ideas what could be faulty, what needs additionally to 
be changed to avoid such messages ?


Also BGP did some messages like that:
Oct 17 04:17:06: [0002]: %BGP-3-ERR: 195.191.170.1 rcv non-connected 
nexthop on ebgp - 195.191.170.5
Oct 17 04:19:00: [0002]: %BGP-3-ERR: 193.33.4.210 rcv non-connected 
nexthop on ebgp - 193.33.4.133

so, a lot of prefixes were dropped down as next hop unavailable, but.. 
after soft-clear they came back without problems..


Regards,
Marcin