Hello,
i use HP dataprotector and VCB on a vsphere 4.0 system, it always worked good but ressently i received error message for some VMs.
here is the error message : "Error: Failed to open the disk: Cannot access a SAN/iSCSI LUN backing this virtual disk. (Hint: If you are using vcbMounter you can use the option "-m nbd" to switch to network based disk access if this is what you want.)"
i targeted the problem to the VCB server, no when i go to server manager and disk i see all my LUN as healty but unknown, thing that is normal, but 1 of the LUN is differend ... the LUN is unknown but not appear Healty and a red circle appears at the place where normaly say the drive is Online ...
i tryied to unplug my FC link then reboot the server but it didnt work ... i dont know what to do the only option i have on this drive is Inisialyze but i dont want to do that ... i am scared to erase data from the LUN in production ...
in attachment you can see an exemple in the Disk manager , note that "SAINS" = healty and "EN LIGNE" = online in french os ....
In the exemple, Disque 5, is a LUN that are ok for vcb backup and Disque 6, is the disk in problem ... always worked but one day that stopped ...
Thanks for answering,
Marc
I still have this problem, is someone got a idea please let me know.
Be sure that LUN masking is the same for all VMFS datastore.
Andre
How can i check that ? in vSphere it seem to be impossible from the UI.
Note that my 3 esx4 server/vcenter can see all lun and the LUN in trouble in my VCB server is fine and usable by those esx.
I mean at storage side.
Andre
yes, all vmfs datastore are in the same group and the zoning is done only once for the group, so the VCB server avec access to the vmfs datastore group that grant access for all the datastore, and only 1 is in problem actually
Thank you Andre,
it was not a masking problem but while checking the masking/zoning i found the problem ... my VCB server only have 1 FC card, so multhipathing is not activated, this card go to a fc switch that are linked to an other switch for redundancy and to the san controlers, but all vmfs datastore was managed by controler A from san and the touble lun was on the controler B, normaly with a multhipathing server with a link in each fc switch its all good, but in this case i needed to have all my vmfs lun on the same controler ! now my problem is solved.
thanks again,
Marc-André