NetXMS Support Forum

Hi Victor,

i have installed v1.2.13 and the on going problem i have with the duplicated nodes still exists.
i said before that i had this problem only on my linux platform but i have 1 node on my windows platform that has the same problem.
this node is based on dhcp and when i changed it's mac address it's ip address changed too. for some reason netxms didn't fixed the values and what i did. is to changed the "primary host name" to the correct ip address and ever since this node is rediscovered. i have to say that i didn't do this procedure on my linux platform but on the linux i have 30 or so duplicated nodes.

we have to find the problem.
Best Regards,
Lindeamon

hi victor,

i do not know what happened but as for now no duplicated nodes were discovered.
i will for a few days to see if the problem is solved

Best Regards,
Lindeamon

Hi!

This is one of the most mysterious bugs I've encounter during NetXMS development Let's wait for a few days to see if problem disappears. In 1.2.12 we found somewhat different problem with subnet objects being duplicated, and fixed this in 1.2.13. Not sure if this is related though.

Best regards,
Victor

hi victor,

good/bad news.
finally all the duplicated nodes are gone except for ~3. i've checked this morning and i found 3 nodes and after deleting all 3 of them only 1 came back again.
last week after i have done the upgrade i have made 2 changes that might have helped to partially solve the problem.
1. i have configured the "active network discovery" to search only x.x.x.x/16 which is my topology instead of no configuration at all.
2. i have disabled "network discovery polling" on a node that might have discovered the duplicated nodes according to the DEBUG log.
3. i have disabled everything related to node name and dns.
just to remind you i have everything on IBM blades and maybe the problem lays there.

what could be the problem?

Best Regards,
Lindeamon

hi victor,

as i mentioned in the previous post, i have disabled "network discovery polling" and that appeared to fix everything. in all this years of using netxms i haven't configured  every node to disable a feature that it does not support like routing or netxms agent. do you think that not configuring might be the problem ?

Best Regards,
Lindeamon

Hi!

Active discovery should not do anything good or bad regarding this problem. I suspect that this is action #2 that helps (but still don't understand why). Do I understand you correctly that you disable discovery polling on only one node? If yes, can you describe this node in as much details as you can (most important are OS, number of interfaces, routing)?

Best regards,
Victor

Hi Victor,

on the 1st time i have done it on a netapp controller. the controller is snmp enabled v8xxx.
it seemed to solve the problem. on the 2nd time i have discovered that apparently i have undid what i have done on the 1st time so i went on and did it on almost every node i have.

hope this helps,
Lindeamon

Hi Victor,

Good News. it has come to a level that i do not get duplicate nodes accept for 1 that i have to check.i haven't changed all the nodes properties regarding the "network discovery poll" configuration and maybe this is the cause for this lonely duplicate node.
something is still bothering me though, if "network discovery poll" shouldn't effect, why the duplicate nodes occured ?
as for your questions, it is a mix of operating systems starting from netapp and snmp support and ending with windows with or without snmp support.
this is very frustrating that i can not give you access to see all the info you need but i think we have to find the cause for this problem.

Best Regards,
Lindeamon

Hi Victor,

i just wanted to run the server in debug to see from where the last node got duplicated and i found another 2 nodes.this time they named "localhost" and they represent the management of the fiber channel modules of the ibm blade center.
if you recall, i have opened a thread on this subject and we didn't finish it at all.
you said that you will check netxms against a blade center, did you do that ? maybe the way a blade center is configured causes problems ?

Best Regards,
Lindeamon

Hi!

It could be because of buggy SNMP agent on blade center for example. Unfortunately I didn't get access to IBM blade center so far. Can you please send me results of SNMP walk on blade center management modules for OIDs .1.3.6.1.2.1.1 and .1.3.6.1.2.1.2?

Best regards,
Victor

Hi Victor,

1st, let me say that after disabling "network discovery poll" to each and every node i do not experience any duplicate nodes.
as for your request, i have attached 1 file  containing the 1st snmp walk and the 2nd string produced nothing.
just to be sure, i have navigated to the desired node, right clicked and chose "MIB Explorer". i have entered the string you wrote and executed right click and walk.

maybe we finally found the problem

Best Regards,
Lindeamon 

Hi!

This dump is for .1.3.6.1.2.1.1 only. Do those devices respond to .1.3.6.1.2.1.2?

Best regards,
Victor

Hi Victor,

no, those devices did not respond to .1.3.6.1.2.1.2
eureka ?

Best Regards,
Lindeamon

hi victor,

problems again. upgraded to v1.2.14 and got duplicate nodes again.
i am starting to believe that it has something to do with the blade center and how it is designed.

Best Regards,
Lindeamon