Node down

On our plant i have total number of 11 nodes. Link of One of Node is down. I attached alarm list. Node 12 is down.. So i plug out NPM and plug in NPM card then link is ok.. My question is that 1st error comes on Node 18 moderror then it comes to other node.. But all other node is ok only node 12 down.

IMG_20190508_072347.jpg

Comments (0) New comment

jabbi_king Rank: 508

Asked 2 years ago

Answer this question Follow

Voted best answer

In Infi90 the loop is one way. So a node down message could be coming from an upstream node. Also loop cabling problems may be associated with this issue. Please contact someone within ABB Service to come out and help with this problem. It is too hard to analyse over this forum.

Comments (0) New comment

Ingo

Rank: 28

Answered 2 years ago

Answers

Hi, A node will only report another node down via the N90STA tag if there are explicit communications (an established route) between itself and the node that went down. Whilst a node forwards all traffic on the ring it is really only interested in communication between itself and other nodes with routes established between them. If the node simply halts due to a fatal error, the timing of any alarming by the other nodes is related to the expiration of any Infinet communications retry processes, so can vary from node to node. From your screenshot, PCU 11 and 14 appeared to have alarmed within 60 seconds of the initial PCU18 alarm, which happened several seconds before PCU 12 was reported as offline.
Always take note of the LEDs on the failed module and associated NIS as these can be invaluable in determining the cause of the module halt. Your specific failure appears to be related to PCU 12 node hardware only as an Infinet failure would have stopped all traffic on the ring with different results.

Cheers,
Alan

Comments (0) New comment

Alan Rank: 13

Answered 2 years ago

Can you please define.. What is NCOMERR and NMODERR. Difference between them when these fault comes?

Comments (3) New comment

by Alan Rank: 13 on 5/14/2019 2:55:46 AM | Like (1) | Report

Hi,
NCOMMERR, NMODERR, NOFFLINE and NSYSERR are summary alarms generated by the N90STA/ModStat tag configured for the node/module. For Comms and Module errors there are numerous causes so a summary error simplifies alarming and graphics implementation. There are 4 categories

NCOMMERR - Summary communications error.
NMODERR - Summary module error.
NOFFLINE - Node Offline error.
NSYSERR - System Error.

In your case...
Node 18 reported a NMODERR, presumably because it would have had remote IO errors for the IO sourced from Node12.
Node 12 and the controller in Node 12 reported as NOFFLINE because Node 12 hardware failed.
Nodes 11 and 14 reported NMODERR probably because they would have had remote IO errors for the IO sourced from Node12.

Refer to 3BUA000157-XXX and 3BUA000158-XXX for the lowdown...

Cheers,
Alan

by jabbi_king Rank: 508 on 5/14/2019 9:42:19 AM | Like (0) | Report

Dear,
Can you suggest from screen shot which node failure exist 1st?

by Alan Rank: 13 on 5/14/2019 6:06:00 PM | Like (1) | Report

Hi,
The guess from the screenshot is that Node 12 failed and Nodes 18, 11, 14 and the connectivity server simply alarmed because they were importing data from Node 12. The actual timing of the alarms doesn't really mean a great deal in this case.
If a node fails it may not have time to issue a node offline broadcast message and the other nodes in the system then have to work this failure out for themselves. As exception reporting is by nature intermittent, it takes time for the other nodes to determine that a node has not sent them any data recently and then go through a process of querying the failed node, retrying and then finally marking the communications with node as failed. Have a look at the NIS/NPM manuals for more details.
Cheers,
Alan

Add new comment

jabbi_king Rank: 508

Answered 2 years ago

Question details

Node down

Voted best answer

Answers

Other questions needing answers

Related topics

AKS experts