How to See Bottleneck(s) on Veeam Restores

Veeam’s bottleneck feature is a great tool for giving a pointer to where a potential problem lies if you are not getting the performance you expect during a backup.
However, when doing a restore Veeam sadly does not show any bottleneck statistics – but it does create them!

To see the bottleneck statistics for a restore is actually quite easy, but does require you know where to look and how to interpret/read them.

Where to Look

Simply go to “C:\ProgramData\Veeam\Backup” and find the folder named after the virtual machine that you are restoring or have restored.
There you will find several log files, but the one that we are interested in here is the one called “Vm.VMNAME.Restore”, VMNAME being the virtual machine name.

What to Look for and How to Read It

The restore log contains a lot of information, but what we are interested in is a line that looks like this (just search for “pex”):

[24.04.2015 15:10:49] <11> Info           [AP] (a82b) output: –pex:0;262144;0;0;262144;1562;79;29;69;79;15;8;130743546495880000

Let us break it down a bit and just look at what is interesting:

  • The number after the sixth semicolon is Source Read Busy at the source storage
  • The number after the seventh semicolon is Source Processing Busy at the source proxy
  • The number after the eight semicolon is Source Write Busy at the source network
  • The number after the ninth semicolon is Target Read Busy at the target network
  • The number after the tenth semicolon is Target Processing Busy at the target proxy
  • The number after the eleventh semicolon is Target Write Busy at the target storage

These numbers are all in percentages and so taking the example above we will get the following bottleneck statistics for the restore:

  • 79 % source storage
  • 29 % source proxy
  • 69 % source network
  • 79 % target network
  • 15 % target proxy
  • 8 % target storage

Now bear in mind that the statistics might change as the jobs goes along, so it is a good idea to go through all the numbers from each line generated by each run. (The percentage processed is the first number after “pex”.)

Leave a Reply

Your email address will not be published. Required fields are marked *