billmathews,

Is it OME 1.0.1 or 1.1 release ? System is down/up alerts are generated internally by OME when discovery/inventory/status poll runs and connection to the manage device is lost/restored. This is not dependent on any protocol used for discovery. May be the connection was lost for few seconds and restored ?

Thanks,

Raj Shresta

This is a fresh install of OME 1.1.  

The majority of our non-Dell devices are still showing Connection Status off.  We have a few switches/non-Dell servers that are Connection Status: on.  But the ones that show off, I can successfully ping from the OME server.  

I've doubled the ICMP timeout (800ms), adjusted to 2 retries, and moved the slider as you suggested.  I'll let that run for awhile and report back.  

Another point to note...

If you believe false system up/down alerts are getting reported, try increasing the ICMP timeout in the discovery wizard and move the status poll slider all the way to the left under Status schedule settings. This will reduce/get rid of false system down alerts.

Thanks !

Nope, still broken.  Every time a status scan runs, it reports a lot of nodes down.  

Thanks for the update Bill.

Are the discovered/classified devices reporting the connection status correctly?

Rob

Thanks,

Ok, anything else unique about these unclassified?  

-Is the only difference that they are unclassified (non-Dell)?

-Subnet or network differences?

-Are they on a WAN or something like that?

Trying to see if there is anything else special or it just boils down to unclassified (SNMP?) devices.

Rob

Hi Rob,

Yes -- the classified devices report their connection status correctly.  It's only the unclassified devices that are a problem.

Bill

Yes, I think that's the only difference.  The devices are on the same subnet as OME and the classified Dell devices.  No WAN or firewall in the middle.  A lot of these are Virtual Machines, but some are physical IBM servers, etc.  

At the moment 2 of my 158 unclassified devices are Connection Status: On (while the rest show Off).  For my classified devices, all 28/28 show Connection Status: On.  

I have this exact same symptom and it;s really annoying!  I have a mixture of VM's and HP servers doing this, all discovered via SNMP and WMI.  Some have a connection status of ON andsome don't.  WHat I have found is that if you do an inventory etc of the subnet/group some of the servers will come back as ON and then over the next few minutes they will return to OFF.