HI All,
Would like to know if there is any possibility to alert for a network device in hung state. As per my knowledge when the network device is in hung state, it is reachable or maybe in unknown state but not sure.
So need your help.
pratikmehta003
Depends on the nature of what happened to cause it to be "hung"
Unless you have done some customizing in NPM if it is still answering pings it is going to show as being up, if not then it will show as down. You have the option to change nodes to get their status from SNMP if you prefer but that has more overhead. If a node shows as unknown it usually means the node has been removed from the database and you need to do some cleanup.
-Marc Netterfield
Loop1 Systems: SolarWinds Training and Professional Services
Hmm so to keep in simple words, it's not possible right..!
And how abt monitoring the service of a load balancer? Is this Something
that can be done by using Pollers?
On 31-Mar-2016 3:31 am, "mesverrum" <
Hey pratikmehta003
If a device supports SNMP, then it is possible that you can create a custom poller for any OID in the SolarWinds MIB database. As mesverrum correctly stated, if you device is responding to ICMP pings, then by default SolarWinds will report the device as 'UP', donated by a green pip next to it's name in the Orion Web Console.
To get anything other than UP/DOWN status from the device, and to create a custom poller to gather information other than compute and interface metrics, you will need to do the following:
The steps above will help you get a custom poller created and assigned to your load balancer. If you need any further information or assistance, please let us know!
Hi silverbacksays
Thanks :-)
Yes that's the understanding i had for ping response for network device and it wont identify as 'hung' state.
I am aware of this Custom poller creation but Network team has given requirements related to service monitoring and all. I am checking on that as well. When i queried the MIB, i found values related to performance and others.
One more thing though, can we have an alert when the SNMP itself is not responding? I have seen in other tools but not sure about Solarwinds.
Can you help me with this?
No problem pratikmehta003!
If you change availability monitoring to track via SNMP rather than ICMP, if SNMP ever fails, your device will be shown as down and will trigger the default 'Node Down' alert. To do this:
There are other ways to do it, but this is the simplest. I wouldn't recommend doing this for a large number of devices, since this would add to traffic on your management network, but for critical kit, it's worth doing.
hmm. i dont think this will work for me. I found few threads about alerting when SNMP is not responding where we can even put the timeslot like only after 5 min or so it should trigger. And the number of nodes i have for this requirement is very less; may by under 5.
And one last thing... when they make any changes on device end, new set of interfaces and volumes will get auto discovered right? We dont have to manually re-discover it.
Unfortunately you will need to discover these new interfaces. However, you can use Network Sonar Discovery to schedule a daily discovery of the effected network segments, or one specific device, it's quite flexible.
On larger networks I advise my clients to run a scheduled daily discovery, just to check if anything has changed. All you need to do is view the discovery results, select the new items, and add them to monitoring. All from within the discovery pages
Then what does the Rediscovery settings do then? In the polling settings we have this option which says rediscovery interval.
Will this not rediscover and identify and later poll and show the status? This is something very basic according to me.
It'll check if anything you have added has changed, but it will not add in new resources, unfortunately.
Thanks silverbacksays
I wasnt aware of this.. its strange though...
