[j-nsp] Memory issues on 10.1R3.7 (J6350)

[Maciej Jan Broniarz]

Hi,

I have a j6350 box, with 2gb ram,  running 10.1R3.7.

For the second time in the last 24 hours my router started to have problems with bgp:

Feb 18 20:40:48  j-6350 snmpd[1007]: SNMPD_HEALTH_MON_INSTANCE: Health Monitor: jroute daemon memory usage (Network time process): new instance detected (variable: sysApplElmtRunMemory.5)
Feb 18 20:40:50  j-6350 snmpd[1007]: SNMPD_HEALTH_MON_INSTANCE: Health Monitor: jkernel daemon memory usage (Network time process): new instance detected (var
iable: sysApplElmtRunMemory.3)
Feb 18 20:42:52  j-6350 rpd[1009]: bgp_hold_timeout:3643: NOTIFICATION sent to XX (External AS XX): code 4 (Hold Timer Expired Error), Reason: holdtime expired for 952.268 (External AS 12968), socket buffer sndcc: 95 rcvcc: 0 TCP state: 4, snd_una: 1099361751 snd_nxt: 1099361808 snd_wnd: 17018 rcv_nxt: 1316583776 rcv_adv: 1316600160, hold timer 0
Feb 18 20:43:43  j-6350 rpd[1009]: bgp_hold_timeout:3643: NOTIFICATION sent to 123.456 (External AS XXYY): code 4 (Hold Timer Expired Error), Reason: holdtime expired for 123.456 (External AS XXYY), socket buffer sndcc: 57 rcvcc: 0 TCP state: 4, snd_una: 1518297759 snd_nxt: 1518297816 snd_wnd: 65000 rcv_nxt: 3880145662 rcv_adv: 3880162046, hold timer 0
Feb 18 20:43:47  j-6350 rpd[1009]: bgp_hold_timeout:3643: NOTIFICATION sent to 567.890 (External AS BBCC): code 4 (Hold Timer Expired Error), Reason: holdtime expired for 567.890 (External AS BBCC), socket buffer sndcc: 57 rcvcc: 0 TCP state: 4, snd_una: 2008592799 snd_nxt: 2008592856 snd_wnd: 66560 rcv_nxt: 186103918 rcv_adv: 186120302, hold timer 0
Feb 18 20:43:53  j-6350 rpd[1009]: bgp_hold_timeout:3643: NOTIFICATION sent to 167.671 (External AS BBCC): code 4 (Hold Timer Expired Error), Reason: holdtime expired for 167.671 (External AS BBCC), socket buffer sndcc: 57 rcvcc: 0 TCP state: 4, snd_una: 251656026 snd_nxt: 251656083 snd_wnd: 66560 rcv_nxt: 709348506 rcv_adv: 709364890, hold timer 0
Feb 18 20:44:27  j-6350 mib2d[1008]: LIBJSNMP_NS_LOG_WARNING: WARNING: AgentX master agent failed to respond to ping.  Attempting to re-register.
Feb 18 20:44:27  j-6350 mib2d[1008]: LIBJSNMP_NS_LOG_INFO: INFO: ns_subagent_open_session: NET-SNMP version 5.3.1 AgentX subagent connected
Feb 18 20:44:32  j-6350 rpd[1009]: bgp_recv: peer 567.890 (External AS BBCC): received unexpected EOF
Feb 18 20:44:39  j-6350 rpd[1009]: bgp_recv: peer 987.655 (External AS BBCC): received unexpected EOFFeb 18 20:44:41  j-6350 rpd[1009]: bgp_process_caps: mismatch NLRI with 13.49 (External AS 1234): peer: <inet-unicast inet-multicast>(3) us: <inet-unicast>(1)
Feb 18 20:45:00  j-6350 cron[82794]: (root) CMD (newsyslog)
Feb 18 20:45:18  j-6350 rpd[1009]: bgp_process_caps: mismatch NLRI with 952.268 (External AS 7234): peer: <inet-unicast inet-multicast>(3) us: <inet-unicast>(1)
Feb 18 20:46:26  j-6350 rpd[1009]: bgp_hold_timeout:3643: NOTIFICATION sent to 25.24 (External AS 987): code 4 (Hold Timer Expired Error), Reason: holdtime expired for 25.24 (External AS 34209), socket buffer sndcc: 57 rcvcc: 0 TCP state: 4, snd_una: 2298471556 snd_nxt: 2298471575 snd_wnd: 16210 rcv_nxt: 820540233 rcv_adv: 820556617, hold timer 0
Feb 18 20:46:33  j-6350 snmpd[1007]: SNMPD_HEALTH_MON_INSTANCE: Health Monitor: jroute daemon memory usage (Management process): new instance detected (variable: sysApplElmtRunMemory.5.6.82787)
Feb 18 20:46:33  j-6350 snmpd[1007]: SNMPD_HEALTH_MON_INSTANCE: Health Monitor: jroute daemon memory usage (Command-line interface): new instance detected (variable: sysApplElmtRunMemory.5.8.82786)

After a few minutes everything went back to normal. 
Memory and CPU usage look fine:

show chassis routing-engine          
Routing Engine status:
    Temperature                 21 degrees C / 69 degrees F
    CPU temperature             43 degrees C / 109 degrees F
    Total memory              2048 MB Max  1126 MB used ( 55 percent)
      Control plane memory    1472 MB Max   707 MB used ( 48 percent)
      Data plane memory        576 MB Max   426 MB used ( 74 percent)
    CPU utilization:
      User                       1 percent
      Real-time threads          9 percent
      Kernel                     0 percent
      Idle                      90 percent


What might be the issue here? Thanks in advance for any help.

All best,
mjb