Community
- Command Central
- MVP Program
- Monthly Mission
- Blogs
- Groups
- Events
- Media Vault
Products
- Observability
- Network Management
- Application Management
- IT Security
- IT Service Management
- System Management
- Database Management
Content Exchange
- SolarWinds Platform
- Server & Application Monitor
- Database Performance Analyzer
- Server Configuration Monitor
- Network Performance Monitor
- Network Configuration Manager
- SQL Sentry
- Web Help Desk
Free Tools & Trials

Network Sonar Discovery Fails - seems to a bug in SW

There was a weird issue with discovery engine process in our environment. Network Sonar Discovery fails for multiple attempts specifically for WMI nodes, although discovery successful for SNMP nodes. We performed Wbemtest for target server with proper credentials which we are using for discovering the servers. Wbemtest works good.

Finally, I had to open a support case in order to isolate the issue. We performed following steps.

1.Re-install JobEngine on both Primary and HA server

2.Repaired Orion core services on both Primary and HA, ran the configuration wizard.

3.Changed polling engine while discovering the nodes

4.Final step we made changes to de-duplication, still there is no luck.

Further, support engineer has reached out Application Dev team to isolate the issue and suggested us to remove all SNMP credentials from discovery profile and run the discovery again. It resolved the issue.

We had around 61 SNMPV1/V2/V3 credentials in discovery profile which leads to time-out of discovery. This looks a bug in SolarWinds Discovery Engine process. Although, we are using polling method via WMI for Windows nodes, Ideally it should bypass the SNMP discovery and run for WMI nodes.

Hope this issue will resolve during next hotfix or version release.

Find more posts tagged with

Status: None

Comments

bharris1

Can you add WMI nodes through the normal "Add Nodes" page? I'd try to add one to see if it can add it successfully.

venubabu

We didn't face any issues adding manually.

mesverrum

If you leave the snmp credentials in the discovery it does not bypass them at all. Before the discovery begins SW does not know what polling method it can use for all the nodes in your list so it walks through every available option one by one. If you are referring to the option where it asks if you would prefer servers to use SNMP or WMI that doesnt not exclude them from trying SNMP credentials, it just asks "if we have a node that answers to both protocols, which should we use?" so it still tests all the snmp credentials until one works before it even begins to test for WMI credentials.

If you want to prevent SW from testing SNMP creds you need to do as the support rep instructed and remove them all completely when you set up the job. This is not likely to get hotfixed out as this is the intended behavior at this time.

Another factor to consider is that you do have some control over the timeout for a job, it defaults to 60 minutes but you can raise it to 10 hours I believe, but if you have so many nodes that 10 hours is not enough to complete the job then you most likely should split them up into smaller batches. I have found over the years that in most environments a discovery that brings back more than 1,000 or so nodes tends to be more likely to have problems completing the import without errors, so I always just start out by breaking the discovery jobs up from the beginning.