Remote backup copy job stuck

[martymidwest]

Hello,

I have a semi-local backup job followed by a backup copy job that goes offsite over a 10M symmetrical T1. I ran my original copy-jobs locally, then took the server offsite. At that time we only had a 5M connection.

Due to the slower original speed, one of the latter listed *large* servers didn't make it offsite. Now, it is almost 2 weeks behind and it cannot complete. Our network seems to have a *blip* about every 8 hours (on average) which is not long enough for the copy job to complete that server. That server is listed 1st now in the copy job, which I though would give it a better chance.

The problem is that every time the network fails for a second, the job restarts from the beginning. It is usually about "35%" which translates to about 650GB transferred.

What can I do to get this VM copied over to that network short of driving the server back over here, wasting several hours of labor and a lot of risk?

WHY does it have to start all the way over every time? Is there any way to make it start where it left off? This is a huge waste of bandwidth, etc.

Thanks,

--mobrien118

[martymidwest]

I guess it's worth mentioning that the error it's throwing is "Error: Client error: An existing connection was forcibly closed by the remote host", in case I'm somehow misreading the symptoms. If I need to open a ticket with support, I will do, but I thought I would reach out to the community first and see if anyone has any easy ideas on how to fix this.

Thanks!!

[Vitaliy S.]

Hello Marty,

Resume functionality is available if you're using WAN accelerators in the backup copy job configuration. If not, then job will start transferring data from scratch after each network drop/disconnect.

Be aware that WAN accelerators were specifically built to address these kind of issues (unstable and slow connection between sites). Can you please tell me if you're using WAN accelerators in your infrastructure?

As to your error message, then it seems related to network connection drops.

Thank you!

[martymidwest]

No, we only have the Enterprise version. We were under the impression when we bought Veeam that our bandwidth would be sufficient for the copy jobs without the WAN accelerators, but I wasn't aware that this level of network stability was to be expected (3 dropped connections in 24 hours really isn't THAT unrealistic, in my opinion).

So, what are our options, other than shelling out for Ent+? Is there any other way to get this job to succeed without moving the server back local temporarily?

Thanks,

--martymidwest

[Vitaliy S.]

Well...basically there are two options, either pre-seed your backup copy jobs or use WAN accelerators to transfer VM data.

[martymidwest]

By pre-seed, you mean that we could make a backup to a local disk and "drive it over" to our DR site and somehow "load" it there? Is this possible given that we already have other backups that have been running for several weeks over either in the same backup repo that we don't want to disturb?

[veremin]

Just seed the job that can't finish data transfer, using procedure described here. In the meantime, the other ("successful") jobs can be left intact. Thanks.

[Vitaliy S.]

You can create a separate backup job, create a full backup file and then follow these guidelines > Recommendations for copying/seeding large backup copy files

Let me know if that helps!