cancel
Showing results for 
Search instead for 
Did you mean: 
Create Post
Level 8

Nodes in Unknown state after unmanage/manage.

Hello,

After putting some nodes in unmanaged mode and putting them back into Managed mode we have a strange behavior: some of them still into "Unknown state" even if we can see status on interfaces.... We have also this strange behavior on ICMP devices...

0 Kudos
22 Replies
Level 13

0 Kudos
Level 7

Having the same issue.... not sure if i'm over utilizing my poller but the SAM application polling Rate is at 72% max rate and total job weight is at 2939. Switching pollers seems to work but yeah that's not a good solution, and I have no intentions of deleting and re-adding.

0 Kudos

Restart the polling service to get them back.

It's what I have been doing for this headache.

0 Kudos
Level 14

This is troubling that this bug has not been acknowledged or taken seriously. I have a client who is beyond mortified that he just lost 6 months of Availability data for all of this clients.

What is the status of this issue?

0 Kudos
Level 13

as far as I know it needs to go to Control Panel -> and select Repair, then run Configuration Wizard again, and it should work

0 Kudos
Level 7

Same issue in NPM 10.7. Any kind of edit to the node, or simply un-manage and re-manage and the node is unknown. Only way I can fix it is to delete the node and re-add it. That method isn't acceptable because once I delete it, I lose all historical data for the node.  I see this going all the way back to 2012 in this thread. Is that the kind of support that this site offers?

0 Kudos

Please follow the post below if you still have the same issue open support ticket .

http://knowledgebase.solarwinds.com/kb/questions/2517/Collector+Data+Processor+and+Collector+Polling...

0 Kudos

I've actually just run into this over the last week or so.  Re-adding the node sets it back to green...but that's not really a answer as it worked fine for years and now it's not.  I have a case open with them...I'll post back if anything comes of it.

0 Kudos
Level 16

I'd open a support case -- I've never seen this in any version up to 10.6.1

0 Kudos

Same issue on NPM 11.  Only way to get them back is to switch pollers.

0 Kudos

That does work sometimes, but recently it has stopped being the remedy that brings them back to life.

   *besides I am tired of changing pollers and rediscovering nodes, when they have not changed

Re-adding the node might resolve the issue , give it a try. Delete and add it again - "but then we would loose historical data"

0 Kudos

This is why we want a solution, that historical data is too valuable to us

0 Kudos
Level 7

Still no answer for this issue?

0 Kudos
Level 9

Bump! (same exact issue), I'm running NPM 10.6.1

0 Kudos
Level 8

Elejeune, did you ever find a solution to this problem?  I am experiencing similar problems.  If anyone else has any ideas, I would like to hear your suggestions too.

0 Kudos

Hi guys,

since the node status is being determined from ICMP echo reply and Interface status via SNMP, there stands several questions:

Are you able to ping the problematic nodes from the Orion server? Are the devices configured to reply on ping and isn't the ICMP port blocked? E.g. by firewall?

Can you, please check it?

Honza.

0 Kudos

Jan, I don't have any issues pinging the servers in question from my Orion server.  Here is a question though.  One of my servers in the unknown state is in the DMZ.  I have to use the FQDN in order to ping that server by name.  Under the Node Details section on the server node page under the DNS field it has the server name not the FQDN.  Can that field be edited?  Does it have any effect on how the ping request is sent out?  How do I ensure that Orion is using the FQDN?

0 Kudos

Hi Iross,

Orion is by default reaching the monitored device via IP address defined during add node or discovery procedure. In case your device's IP is dynamic (assigned by DHCP, etc.), or you need to define the hostname (in your case full FQDN), go to the Node details or Node management and go to edit node section. Then under the Polling IP address text field, check the "Dynamic IP Address (DHCP or BOOTP)" checkbox.

Then you can define directly the hostname (FQDN), which will be used for reaching the device (DNS resolution must work).

Please let me know if it helped.

THX.

Honza

0 Kudos

Thanks for the info Jan,

I wasn't sure if Orion used IP or System Name to ping for up down.  We are using static IP addresses on our servers, so that shouldn't be an issue then.  As I said I can ping my unknown status server from the Orion server both by FQDN and IP address.  I even have been getting alerts on Windows Server Counters that have gone into the warning state.  Everything on this particular server is green except for the Node Status which is unknown.  Any other ideas?

0 Kudos