After the NPM module crashed, lines similar to the ASSERT (shown in red) can be found in the messages:
Sep 17 12:58:41 npm1 cbsnpmcfgd[551]: [W] [npm1 1.1.1.1] Lost contact with Config Manager.
Sep 17 12:58:41 npm1 cbsnpmcfgd[551]: [W] [npm1 1.1.1.1] ... attempting to restart connection to Config Manager.
Sep 17 12:58:42 npm1 cbsnpmcfgd[551]: [I] [npm1 1.1.1.1] PhysIntfDrvrMon cbsif monitor connected to cfgmgr on primary CPM (1.1.1.20:9720)...
Sep 17 12:58:42 npm1 cbsnpmcfgd[551]: [I] [npm1 1.1.1.1] Config Agent top-tier connected to cfgmgr on primary CPM (1.1.1.20:9720)...
...
Sep 17 12:58:42 npm1 cbsnpmcfgd[551]: [I] [npm1 1.1.1.1] PhysIntfDrvrMon detected lost connection to CPM; waiting to re-connect...
Sep 17 12:58:42 npm1 cbsnpmcfgd[551]: [I] [npm1 1.1.1.1] Flushing NpmFpmStateTable...
Sep 17 12:58:42 npm1 cbsnpmcfgd[551]: [I] [npm1 1.1.1.1] Flushing NpmXrrpVrGroupStateTable...
Sep 17 12:58:42 npm1 cbsnpmcfgd[551]: [W] [npm1 1.1.1.1] Could not insert synthesized CircuitTable Entry.
Sep 17 12:58:42 npm1 root: cbsnpmcfgd: [F] ASSERT failure expr="oldEntry != __null", file=npmEzL2PolicyManager.cpp, line=1139
Sep 17 12:58:58 npm1 cbsinitd[763]: [W] [npm1 1.1.1.1] daemon 'daemon.cbsnpmcfgd' pid '551' not running (3:No such process)
Sep 17 12:58:58 npm1 cbsinitd[763]: [I] [npm1 1.1.1.1] killing any remaining running pids for daemon 'daemon.cbsnpmcfgd' (prcnam:cbsnpmcfgd)
Sep 17 12:58:58 npm1 cbsinitd[763]: [I] [npm1 1.1.1.1] starting daemon 'daemon.cbsnpmcfgd' using command '/sbin/reboot -d0'
Sep 17 12:58:58 npm1 init: The system is going down NOW !
This issue may occur during an XOS upgrade and should not have any impact during normal operation.
The issue has been fixed in XOS 10.0 and later.