I'd log a case with vmware support. Just a thought, is the storage servicing your vCentre swamped?
Hooo yeah... lot of strange stuff...
To define scope :
One cluster HA/DRS with 6 nodes, 14*700GB DS, 98 VM
One datastore has on 340MB free... on 700GB, but status is "normal"... Okaaaayyy..
I've started a global vcheckv505+, it returns free space problem (fortunately!).
On this cluster we've a pb on a datastore which has been corrupt, and we cannot remove it on 2 ESX... we need to reboot, but timeline not still defined (ressource are too short too migrate vm on other ESX...)
Some path failed... but not much considering number of links...
Disk latency : high on 2 ESX (concerning DS which cannot be removed without reboot...)
small on some others esx (53Ms on 30 defined)
2 vm ballonning... we'll correct that.
Concerning this DS with 350MB free, i've moved a VM : display value was actualized, so i'll migrate VM to have a preconised free space... but i don't know why triggering doesn't alerting... (configuration of triggerring is correctly defined on VC)
Needs others specific information to answer your doubt?
Some specific logs ?
command to test?
Thanks for your hlp
So what is the datastore latency showing, on the vCentre VM?
Vcenter is on a standalone server...
problem solved on datastore free space : all of them have 20% free...
alerting triggers problem has been solved too : change values n return to previous... warnings and alerts appaered..
no more idea about graph problem? :s
Was your vCenter build from scrach or you did upgrade to 4.1 ? I had similar problem but with historical data not with realtime.
Does error appears on all hosts in your cluster ?
Where is vCenter DB located, is it on this same server where vCenter is installed or remote DB ?
What version of DB engine is used for vCenter DB ?
Vsphere has been installed from scratch... no update.
Problem appears on different nodes... but not at the same frequence...
Vcenter DB is located on another SQL server... SQL 2005 I think... (SQL Native client 2005.90.30.42.00)
That's odd, looks like a problem between ESXi hosts and vCetner.
have tryide restart mgmt services on hosts ?
Have you tried disconnect and connect host again - to renew vpxd configuration ? remove and add Host to vCetner inventory ?
check SQL performance itself
What are statisctics levels on your vCenter ?
Is there anything suspicious in vpxd*.log on vCenter server ?