Node missing in an Advanced Network of 8 panels ????

[Benzerari]

Hi all;

One of our customers has a large building of 7 floors Inside (NHS), with one advanced system in each floor, all networked together...

If alarm goes off in one of the floors it sets intermittent alarm in the floor above and below... mean while it sends a signal to switch board computer, stating the input triggered with the time and date...

The bad incident is that one of the mental patients had broken the glass in the 2nd floor then alarm went off normally, but it did not set intermittent alarm in the floor above and below.

When checked all panels I found the one in the first floor was hidden inside a cupboard it was showing 'Next Node Missing' which is found to be the 2nd floor panel it was behaving as stand alone panel and displaying 'Node 0'

I have reprogrammed the nodes back to normal but the customer's question that has really struggled me, and I could not find any answer to it is  :/

What makes the node to go missing?

Thank you

[Graeme]

is the network fault tollerant or standard?

it should be fault tollerant but if it's standard then is the panel last in line?

The panel should not be able to revert back to node 0. Bad packets,faulty network card or fault in cable could cause a fault on the network and make the panel stand alone.

[David Rooney]

I'm not that familiar with Advanced kit but surely if a network card address disappears off the system then surely this should be flagged up on all the other panels ??

I know their fault tolerant cards are a lot of money... perhaps that's the answer......

[Benzerari]

Thanks graem and david for the reply;

Only the main panel ( the master one ) in the ground floor was displaying 'fault in the first floor', when checked the panel in the first floor, it was showing 'next node missing', but all the rest of the panels were totally healthy.

My concern is more about, why the panel of the second floor went back to 'node 0' instead of 'node 6' as it was originally?

The senario I have been given, has no logical link to the node missing incident, that some mental broke the glass, and the tenant has managed to repair it....    even the code provided to the tenant would not allow access to the settings ( level 3 ) and make some changes...

It must be some serious cause... the customer insisted into the fact that the node has gone missing must have a cause behind it... which is fairly logic...

I still do not know

[kurnal]

As a non technical user of alarm systems - I drive them and dont try to fix them-  could I ask if the panel or card has lost its configuration due to a fault and has reverted to basic configuration stored in onboard memory - I have seen systems do this to ensure that if the whole network falls over at least the node will still be able to control its own devices but in isolation from the rest of the network. So network causes, effects  and configuration are lost as it reverts to rock bottom basic operation firing off every interface for every cause.

[Benzerari]

As a non technical user of alarm systems - I drive them and dont try to fix them-  could I ask if the panel or card has lost its configuration due to a fault and has reverted to basic configuration stored in onboard memory - I have seen systems do this to ensure that if the whole network falls over at least the node will still be able to control its own devices but in isolation from the rest of the network. So network causes, effects  and configuration are lost as it reverts to rock bottom basic operation firing off every interface for every cause.

The panel of the second floor lost its config. within the network only. But its basic config. had not been lost I had checked the device log it was ok.... the fact that when that mental broke the glass, alarm went off in that floor, but no intermittent alarm in the first and third as it should be....

I will e-mail advanced electronics ltd technical support, for any bit of light for future measures...

[kurnal]

So in my laymens terms the 2 stage alarm is configured on whichever panel controls the network and this normally over rides local settings on the other nodes, but if a node loses contact with the network it reverts to its own cause and effect settings in accodance with its own configuration. I have seen this before. Its all down to the configuration matrices and whether network commands over ride local commands.

If network commands over ride local commands, when the system is commissioned all seems ok, but if ever the network fails or loses its config, the individual nodes will revert to their local config.
Had an IT suite unexpectedly closed down once by this happening. Cost the client a fortune.

[monkeh]

was there nothing in the event log to suggest someone had entered the panel and adjusted settings?


if not then there must be some fault with the network card that has caused it to reset.

[Graeme]

let us know if you get a reply from Advanced.

[Benzerari]

let us know if you get a reply from Advanced.

I have e-mailed them on Friday afternoon, hopefully by Monday morning I will get some answers.

[Wiz]

In this scenario, I am assuming that the sounders on floor that didn't pulse are connected to one networked panel but the MCP initiating the alarm was connected to a different networked  panel.

Whilst most networked systems,  that I have experience of, operate by having the same operation software configuration stored in each network node of the system, if you lose the connection between panels, cause and effect functions initiating on one panel but requiring an operation on another panel, might not work.

The simple fact is that if you lose the connection between two panels (nodes) then the signal initiated from the first panel cannot be seen by the second panel. Therefore, even if the programme in the second floor was configured for pulsed alarms on receipt of an alarm condition from a device on another network node, if it doesn't receive the signal then it can't initiate it's function programmes because it doesn't even know an alarm exists.

Therefore the panels haven't reverted to a different configuration because the network connection has failed, but simply because the initiating signal hasn't got through.

Networks should be a loop and capable of transmitting in both directions around the loop to avoid a simple disconnection at one point causing a failure of operation

In this scenario, the loss of the network in any form should have been indicated on all other networked panels as a fault and the user should have been aware that this was a serious fault that could have serious consequences to the operation of the alarm.

If no fault was showing, it must be that the configuration programmes in all other working panels was incorrect and not even looking for the missing 'network node'.

[monkeh]

In this scenario, I am assuming that the sounders on floor that didn't pulse are connected to one networked panel but the MCP initiating the alarm was connected to a different networked  panel.

Whilst most networked systems,  that I have experience of, operate by having the same operation software configuration stored in each network node of the system, if you lose the connection between panels, cause and effect functions initiating on one panel but requiring an operation on another panel, might not work.

The simple fact is that if you lose the connection between two panels (nodes) then the signal initiated from the first panel cannot be seen by the second panel. Therefore, even if the programme in the second floor was configured for pulsed alarms on receipt of an alarm condition from a device on another network node, if it doesn't receive the signal then it can't initiate it's function programmes because it doesn't even know an alarm exists.

Therefore the panels haven't reverted to a different configuration because the network connection has failed, but simply because the initiating signal hasn't got through.

Networks should be a loop and capable of transmitting in both directions around the loop to avoid a simple disconnection at one point causing a failure of operation

In this scenario, the loss of the network in any form should have been indicated on all other networked panels as a fault and the user should have been aware that this was a serious fault that could have serious consequences to the operation of the alarm.

If no fault was showing, it must be that the configuration programmes in all other working panels was incorrect and not even looking for the missing 'network node'.

our site protfolio only includes one set of advanced panels networked together, and they are simply two panels side by side because the site is too big for one panel to cover it, but in that setup, and with morley panels (which, lets be honest, are virtually the same thing), only the master panel of the network monitors the others.


in any of our sites that have various panels networked together, it's all monitored by the drax amx software. even the sites we have that include morley panels networked, all the seperate panels are either indepentantly networed, or the main control panel is networked, meaning that any fault with the comunications is flagged up immediately with the responsible parties.

in  this case, it shows that the master panel of the network being boxed up in a cupboard is a bit of a daft idea.

[Benzerari]

In this scenario, I am assuming that the sounders on floor that didn't pulse are connected to one networked panel but the MCP initiating the alarm was connected to a different networked  panel.

This is not in our case, both sounders and call points are part of the same and unique panel of that floor

Networks should be a loop and capable of transmitting in both directions around the loop to avoid a simple disconnection at one point causing a failure of operation
In this scenario, the loss of the network in any form should have been indicated on all other networked panels as a fault and the user should have been aware that this was a serious fault that could have serious consequences to the operation of the alarm.

Also this is not our case the fault was dispalyed in the master panel ( Gd Floor one ) stating that 'fault on panel of the first floor', and the one of the first floor displays next node missing which is the second floor one. all the rest were healthy. I think it depends of how you program the faults to be reported to the main panel or to all panels.

If no fault was showing, it must be that the configuration programmes in all other working panels was incorrect and not even looking for the missing 'network node'.

The other panels did not show 'next node missing' because none of them was missing and each one of them seen or monitored by the previous one.

Our customer is hundreds of miles away from us, by next due visit, I will try to download all required details, 'Event log, Softwares versions of each panel...etc'

[Graeme]

The other panels would have to have been showing a network fault though?

with Advanced networking with say 5 nodes. node one looks for node 2  2 for 3  3 for 4  4 for 5   and 5 for 1.

so if any of these nodes went mising then a network fault would be flagged up.

[Graeme]

out of interest i take it 7 floors and the 2nd floor was node 6 then the network cannot have been wired as node 1 to 7 as recommended?

i would have had node 1 ground floor  2 1st floor   3 2nd floor etc..

but for node 6 to be on the 2nd floor would suggest that the wiring does not go from node 1-2 etc and it could be a case that node 6 has never been wired in to the network and it sits last in line and the other panels which are currently programmed as a seven panel network will be nother the wiser as it's  a radial network and not fault tollerant.

the stand alone panel if not wired into the newtwork correctly and programmed to node 8 will be looking for node 9 or 1 which will cause node missing on this panel only and not a network fault globally.