Slow throughput after completing first disk of a vm

[KeiichiKun]

Hi,
we have a backup copy job (Veeam 9.5) to a dr remote site over a 100mbit connection (I know it's slow for 3tb data, but that's what we have).
Step to reproduce the problem:
- start backup copy job, it starts at 11 MB/s (first disk vm1) + 300 KB/s (first disk vm2) + 250 KB/s (first disk vm3)
- analyze ping between veeam vm on production site and remote vm on dr site -> ok
- wait for completion of first disk of first vm
- second disk of vm1 starts at 200KB/s
- ping goes down and we have lost packet

During last two points, bandwith is not full and no other jobs are taking free band

If we stop and restart the job, 11Mb/s is used again, until first disk ends.

Case id #01982157

Thanks

[Crof]

Out of curiosity, is the ping delay increase also seen between two machines in your sites that are not involved with the backups? Does the ping delay happen when pinging an outside servers like googles DNS ( 8.8.8.8 )?

[KeiichiKun]

Nice question.

- ping from veeam vm on production site to vm on remote dr is slow
- ping from veeam vm on production site to 8.8.8. is fast
- ping from another vm on production site to vm on remote dr using the same 100mbit (this line is dedicated to backup) is fast
- ping from veeam vm on production site to another machine on production site is fast

It seems that veeam can't understand that line is not full after first disk completion..

[KeiichiKun]

An important information.
If I start a job on another veeam instance (9.5) on another production server and another repository for backup job, but using the same remote vm and same line, speed is slow!

[KeiichiKun]

There was a network error on friday that corrupt something on fc port on core switch and on gbe port connecting remote dr.
After restarting the storage and the gbe port everything seems working fine.

[Crof]

So it was a networking hardware issue at the remote site?

Glad you found a fix