Looking for advice about low restore throughput

[andrefaria]

Hello, I'm driving some tests for restore from Exagrid to Pure storage, my Infrastructure is the following:

Explaining this image:
Pure storage are connected at 8Gbps to 2 brocades 8Gbps and them they are connect to host with 8Gbps HBAs.
The host 10 GbE NIC are connected to 2 Cisco 6125 that therefor are connected to Exagrid with a LACP with 2 10GbE FC.

My VBR VM hardware is 12 vCPU / 12 GB RAM / VMXNET3 and it sits inside the same ESXi Host that I'm restoring the virtual machine, I'm testing with a 2 disks VM, 300 GB and 33 GB, as result to my tests I've got the following:

At begin of the restore my network stats 1.2 Gbps for reading from Exagrid and ~150MB/s for writing to Pure storage, after restoring the smaller disk it drop to approximately 650Mbps for reading from Exagrid and 70MB/s for writing to Pure storage.

[foggy]

Hi André, what transport mode is being utilized during restore? You can find it logged as a [tag] after the proxy server name in the restore job session. Also, our support will be able to tell the bottleneck by reviewing the restore logs.

[andrefaria]

As per I see, it's [hotadd].

I didn't thought that was the case to open a ticket but I will follow your advice, thanks.

[foggy]

So seems everything is configured correctly. Let's see what our engineers say after a closer look.

[andrefaria]

FYI

CASE #04710537
Thanks again.