Fault node on network startup

[alldone]

When we woke this morning (in New Zealand so it's winter) our radiators were cold. As far as I can tell, the cortex PC restarted at 3.20 am, probably due to an automatic windows update, and one node (our outside LTH) faulted. The log message is "some nodes failed to reset" and led to a "network stop request". This seems to have caused the network not to start with the fault screen/node status screen displayed. When restarted, the node reset normally and everything was fine, but the house was a bit cold.

Sometimes a node will stop working in our system for no particular reason and when reset it is fine. The "fault node and stop communicating with it" setting in the network supervisor is set but doesn't seem to apply at network start (since the network doesn't start). The autoreflex program for the heating seemed to work when I tested it some time ago when I powered down the PC, but having partly started the network, the reflex programming seems not to have kicked in.

The IPD repower spur on fault was set to only operate between 0800 and 2200 (the default I think), though I have now changed this to "all the time." Will this then repower/reset the faulty node so the network can start - I suspect it probably wont and am looking for a way to get the network to automatically always start, so we don't lose all our heating in the middle of the night again.

Any ideas anyone?

[Viv]

In Network Supervisor | On Network Run | set 'PC to Network Interface' and 'Network nodes' to "No reset".

Resetting nodes is not really necessary and was introduced before the IPS was developed.

If a fault develops during the network start phase and the network is set to 'Run on program start' then it will progress with faulty nodes to get as much of the network running.
If it was user started, then the user is offered a choice as to whether to proceed or stop.

Their may be no logical reason why, even with the reset option turned on, that it could not proceed as above and we will investigate to see if a Cortex update is suitable.

Viv

[alldone]

Thanks Viv. I had wondered about changing these to no reset but figured that resetting the system would be safer. Obviously a very rare combination of circumstances as I find my cortex installation is very stable in general (usually only windows update restarts and power cuts cause a restart) and it is rare to have a module fault. I have made the changes you suggest.

[Karam]

Nonetheless, it is a good idea to try and find the cause of any node faults, so if you can remember to do so - when you experience a fault it would be useful if you could send us any log files captured around the time of the event (found in ....Cortex/log folder) with a description of any other symptoms (include module type, name in database and ID number). Check that your General logging options are: Enabled, Autolog on failure to File, Time, Network traffic, Object names. This is generally the default.