RNRP errors in AC 800M controller logg...
I get the following errors in one of my AC 800M controllers after firmware upgrade to latest version build 220.127.116.11.
E 2015-11-05 17:23:44.426 RNRP: SystErrLog=631, errNo=-1408217063 param=03000000
E 2015-11-05 17:23:44.426 RNRP: SystErrLog=631, errNo=-1408217063 param=03010000
W 2015-11-05 17:23:54.336 Unit= _SW1131Task Controller_3 2002 Interval time in ordinary tasks dec 1.3
W 2015-11-05 17:24:44.736 Unit= _SW1131Task Controller_3 2001 Interval time in ordinary tasks inc 1.3
W 2015-11-05 17:24:54.836 Unit= _SW1131Task Controller_3 2002 Interval time in ordinary tasks dec 1.3
E 2015-11-05 17:25:43.896 RNRP: SystErrLog=631, errNo=-1408217064 param=01000000
E 2015-11-05 17:25:43.896 RNRP: SystErrLog=631, errNo=-1408217064 param=01010000
The CPU is of modell PM861 and we use redundant networks.
Anyone know the meaning of the RNRP errors?
Because of the RNRP error the CPU sometimes go to almost 100% load and as you see on warnings trying to change task times.
Thank you in advance.
Best Regards // Mattias
The SystErrLog=631 is emitted when RNRP in an AC 800M has detected loss of a router node to a remote RNRP area (e.g. a connectivity server, in this case convert the negative errNo to hex and look at the eight lowest bytes -1408217063 = FFFFFFFFAC105019 = 172.16.80.25) and forces it to find a new router. If no other router is available, this error is output.
If I got it correct, "631, -1408217063, 03000000" = I was not able to find an alternate router to area 3 when 172.16.80.25 was lost. Sorry.
The increase in CPU load is probably originating from some application using Ethernet network communication (MMS, IAC, ModBusTCP, etc).
The controller load can take very large jumps when network connections are established or disconnected if you have numerous connect/read/write blocks and/or quickly executing tasks.
I'm no real pro on AC 800M programming, but I've seen recommendations to try to stagger (spread over time) the connect/read/write blocks to prevent overload. I.e. prevent such blocks from reacting on network events all at the same time.