Saturday, April 28, 2018

PowerCLI - Batch migrate VM network adapter

Back to VMware basic operational tasks: I had to migrate VMs from one network to another, which roughly translates to changing the portgoups Pretty simple and straightforward. Since the task was for several VMs, I automatically excluded clicking through the UI. That sent me to PowerCLI and now the simple task became a bit more complicated because instead on relying in real time on my hand-eye coordination for changing a portgroup, I would have to rely on an input file.

The input file is in CSV format and has four columns: VM name, source portgroup, destination portgroup and reboot (boolean value to see if reboot the VM or not). You may ask why using the source portgroup as input - I am using it to check that the VM actually has one network adapter connected to that portgroup and I am not randomly migrating everything I find on that VM.

The CSV file looks like this:
vmName,srcPg,dstPg,reboot
vm-1abc*,pg-prod-101,pg-prod-110,false
vm-2def*,pg-prod-101,pg-prod-110,false

I am also using wildcards in the name of the VMs. The reason behind this is the VMs have a very long and randomly generated name which is different from the hostname. For example, actual VM name is vm-1abc-yetcg-93763-andbv-34781, while hostname is vm-1abc. How I am checking that the wildcard does not match multiple VMs, you will see in the description below.

The migration script does the following:
  1. takes as input the CSV file and tries to load and read the CSV file; if it fails, it will exit the execution
  2. for each line in the CSV file it searches the VM in vCenter server inventory
  3. once the VM has been uniquely identified, it searches for a network adapter connected to source portgroup as defined in input file; it doesn't find it or finds multiple portgroups, it does not process the VM
  4. once the source portgroup is identified it changes the network adapter mapping to destination portgroup (as defined in input file)
  5. if reboot is required, it will issue a soft reboot to the VM
  6. and it goes to the next VM in the list

And now the code: as usual, please use it carefully as it has not been tested for all situations. Also, PowerCLI session from where the script will be run needs to be connected to vCenter Server as the script does not handle this. 



 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
param(
 [Parameter(Mandatory=$true)][string]$csvVmList
)

function VmChangeNetwork($vmName,$srcPg,$dstPg,$reboot){
  Write-Host "processing:" $vmName
  Write-Host "  src PG:" $srcPg "dst PG:" $dstPg "reboot required:" $reboot

  Try {
      $v = Get-VM | Where {$_.Name -like "$vmName"}
  }
  Catch {
    Write-Host $_.Exception.Message $_.Exception.ItemName
  }

  if ($v.Count -eq 1){
    $srcPgExist = $v  | Get-NetworkAdapter | Where {$_.NetworkName -eq $srcPg}
    if ($srcPgExist.Count -eq 1) {
      $v  | Get-NetworkAdapter | Where {$_.NetworkName -eq $srcPg} | Set-NetworkAdapter -NetworkName $dstPg -Confirm:$false
      if (($reboot.ToLower() -match "true") -and ($v.PowerState -match "PoweredOn")){
        Write-Host " rebooting VM"
        Restart-VM -VM $v -RunAsync -Confirm $False
      }
    } elseif ($srcPgExist.Count -eq 0) {
      Write-Host " no adapters connected to" $srcPg "found"
    } else {
      Write-Host " multiple adapters connected to" $srcPg "found"
    }


  } elseif ($v.Count -eq 0) {
    Write-Host " "$vmName "was not found"
  } else {
    Write-Host " "$v.Count "VMs found with name" $vmName
  }
  Write-Host ""
}

# load CSV file
Try {
  $vmList = Import-Csv $csvVmList
}
Catch {
  Write-Host -foregroundcolor red " File is not accessible"
 exit
}

# process VMs
foreach ($vm in $vmList){
  VmChangeNetwork -vmName $vm.vmName -srcPg $vm.srcPg -dstPg $vm.dstPg -reboot $vm.reboot
}

Wednesday, April 11, 2018

Veeam Backup and Replication - Infinidat Integration

Starting with Update 3 (U3), Veeam Backup & Replication (VBR) offers a built-in integration framework for storage systems called Universal Storage Integration API. Storage vendors can use the API to develop plugins and integrate their storage systems with VBR. This is a huge step into extending the ecosystem of storage vendors that offer advanced functionality with VBR.

As of the writing of this article,  the following storage systems are already supported:
  • IBM Spectrum Virtualize (since December)
  • INFINIDAT InfiniBox (since March)
  • Pure Storage FlashArray (since April) 
(Later edit: first storage vendor supported via the API was IBM, but it is integrated in U3 release)
The following article presents the integration with Infinidat InfiniBox.

I will not go into the details of installation and configuration of InfiniBox. The post covers only Veeam part

First, download the Infinidat plugin from Veeam site. Login to VBR server, close VBR console, make sure processes/jobs are not running, extract the zipped file and run the installer. If you connect remotely to VBR, then run the installer on the machines from where you connect, too. It's a next, next, next process (as seen in the following series of pictures):





Now, getting to the fun part. Open VBR console, go to Storage Infrastructure > Add Storage - there it is, Infinidat Infinibox is available:



Add storage hostname (or IP address):

Add storage credentials:

Select the protocol to use (in lab I am using only iSCSI), the volumes to scan and the proxies to use (I have left everything on automatic):

Review the summary, press Finish and wait for the successful installation:


It is time to test the newly configured storage. First, let's create a job that uses Backup from Storage Snapshots. Since using integrated storage snapshots is active by default, there is no need to configure anything special for this job. Just select the VMs from InfiniBox datastore and run the job. Looking at the logs we see that "storage snapshot" is being used for backup:

Another type of job is to use only snapshots, although this is not a proper backup solution (since both the source VM and the backup reside on the same storage). To make a backup job snapshot only, you need to select as destination repository the InfiniBox storage (not a VBR repository):

This time when the job runs, it will create a snapshot directly on the storage:

The snapshots will appear under Home - Backups:

As well as under Storage Infrastructure:

Now, it's time for you to test the recovery :-)

Wednesday, April 4, 2018

Veeam ONE Custom Reports

I was recently asked if Veeam ONE could also create a custom reports, such as an inventory list of the VMs and their configuration.

Veeam ONE does come pre-loaded with a lot of reports. But if none of those reports are satisfying, then you can create your own.

To do this, first we login to Veeam ONE Reporter and go to Workspace tab. In the left pane, under My Reports, we create a folder (not mandatory, but a good practice to keep things separated).


Next we see a preexisting folder, Custom reports. We open folder and select "Custom Infrastructure" report which allows to define our own parameters.


The selected scope is Virtual Infrastructure and allows to select from all the objects. We are interested in vSphere Virtual Machine, but it could be any vSphere (or Hyper-V) inventory object.


Next we select the parameters to display, for our case: VM name, number of vcpus, memory size, disk size and IP address. The window allows for real time filtering of the available parameters (makes life easier than scrolling through a long list):


In case we are looking for something specific (let's say VMs that have the letters "vbr" in their name and less than 32 GB of RAM) we can actually create a custom filter and only those VMs will be displayed.


Cool, right? To make it even more flexible we can group the VMs based on one of the properties (for example memory size) and we can also choose to sort the lists. After we finished tweaking with the report, it's time to save it (to the folder created earlier).


After the report is saved, we go to the location of the report and run it (of course it can be edited, copied, deleted or scheduled to run periodically and sent as attachment to an e-mail address)


And the result of running the report is:


This is just an example on how to use custom reports. The power of Veeam ONE comes from letting you choose any parameter from the monitored infrastructure (virtual and backup) and use it in your own custom report.