Hi,
We have around 8 ESX servers running in a cluster.The VMs in this farm are getting frozen frequently.The only way to bring the VMs back to running state is to kill the PID.when i checked the hostd.log i see the below mentioned error.
VmfsAIOMgrOpen: Device or resource busy
DISKLIB-VMFS : "/vmfs/***********.vmdk" : failed to open (1049095): AIOMgr_Open failed..
DISKLIB-DSCPTR: : "/vmfs/***********.vmdk" : Failed to open vmfs disk.
"/vmfs/volume/virtual machine.vmdk" : failed to open (1049095). Descriptor file couldn't be read
"/vmfs/volume/virtual machine.vmdk" : failed to open (1049095).
"/vmfs/volume/virtual machine.vmdk" : failed to open with flags 23 (1049095).
Any idea what could be the reason.The VMs are located on FC SAN datastores.
:
Is there high IO load on datastore? If so try to find most IO loaded VMs and put them to separate datastores located on separate disk groups.
---
MCSA, MCTS, VCP, VMware vExpert '2009
Thanks For your response.How can analyze the I/O load on a datastore.Is there a Tool or document on how to do analyze the load on the Datastore
Thanks,
Shanu
Veeam Monitor
---
MCSA, MCTS, VCP, VMware vExpert '2009