Hi Veritas support.
First of all, i'm a total Netbackup beginner, please forgive my ignorance. I'm opening this case today because of an issue i've seen described already on this support community, but for NetBackup 7.7. ( most likely 7.5 )
The context :
I'm currtently using Netbackup 7.7.2, in order to backup virtual machines from a specific ESXi 5.5.U2 host. This host is not integrated with any VCenter , and no DNS server is used on the 1Gb network between the ESXi host, and the NetBackup server ( resolution is done by hosts files )
I've a physical server (W2012R2 ) as Netbackup master server. This server has a SAS HP Tape Autoloader loader attached to it. 1 driver 8 cartdriges. It is also integrated with another iSCSI VTL, emulated by a HP StoreOnce 2700.
The virtual mahcines were added to the polocy, by manual selection.
The problem :
I built a very simple policy, which backups the 2 virtual machines, with one full once a week at 00:00 and then 1 inc every hour. (See attached file ). so far all the VMware optimization are enabled in the VMware tab of this policy. The schedules do not really make sense, but this is just for testing so far.
Whenever i do a full backup, everything goes fine, and the full backup is processed at ~ 60Mbytes/s for the 2VMs. After this, if i try to execute an incremental backup, the activity monitor tells me that the DIFF Job is starting, but i actually observe that as much data as the full backup is pumped out of the ESXi to the backup server. The diff takes as long as the full, and looks as big at the end.
Could you please help me to figure this ?
Solved! Go to Solution.
It turned out to be generated by a version of ESXi 5.5 U2 impacted by an issue in the CBT feature.
Upgrade to the latest ESXi 5.5 U3 , plus CBT parameters reset / ctk files cleanup in the ESX datastore fixed it.