Big drive backup over slow connection

[FireExit]

Hi,
This server has 1.65TB of D:\ drive and 14Mbit/sec internet upload.
It's doing a VAW volume backup - after 5 days, it's 12% done.
At this rate it will be a month, but I doubt it will run uninterrupted for a month - patching updates if nothing else.

Is there any possible workaround or helper?
Any way to do a volume backup which is restartable, e.g. using a Windows Shadow Copy point?
Any way to backup to a local NAS and then copy to Cloud Connect (S3) without using B&R?
Any way to get VAW to talk to one of their laptops and piggy-back a nearby cafe wifi

I have considered splitting it up into smaller drives, but there's one big data folder with about half the data; might help but not an easy solution.
I have considered ramping up the compression, but default "optimal" compression is pretty good already - and a step change of even 250GB doesn't significantly change the situation.

I think they previously had a seeded backup locally and moved off-site, but with CloudConnect backed by S3 object storage I can't see that being an option.
Ridiculous request, they need faster internet... but anything big I've missed?

[Nils]

That ratio of backup size and upload rate is likely not practical. I'd recommend using local storage (or locally attached network storage) and then either copying that to the remote location or moving a local storage device physically off premise.

Separating backup and remote copy tasks has the upsides of a) backup finishes within a reasonable backup windows and b) a copy interruption can easily be mitigated using the proper copying mechanisms. A local backup is also the only way to enable a restore within a reasonable time window with that kind if Internet connection.

[FireExit]

Thanks Nils, that's about what I expected. :/

> I'd recommend using local storage (or locally attached network storage) and then either copying that to the remote location

1) it's a Veeam Agent for Windows backup, and that doesn't seem to have an option to do a local copy then do a secondary destination copy job to a remote location(?)

2) if they don't have the bandwidth to do a copy to a remote location, then what changes if it's staged locally first?

> or moving a local storage device physically off premise.

I don't know of a way to do this when the target is S3/Wasabi cloud object storage(?)

[Nils]

> Veeam Agent for Windows [...] doesn't seem to have an option to do a local copy then do a secondary destination copy job to a remote location(?)

No, it doesn't. You need to use another tool for replicating the local backup to a remote destination. robocopy comes to mind, or sneakernetting a USB drive, or ...

I've created a script for my local backup (to SSD) that syncs that backup to cloud and creates another, HDD copy once the Veeam job is done. A simple

Code: Select all

:waitforveeamstart
timeout >nul /t 10
tasklist | find >nul /i "VeeamAgent.exe" || goto waitforveeamstart

does the trick.

The point is to use a tool that allows you to efficiently resume an interrupted copy process.

I was replying in general, not specifically for S3 storage. Likely, there's a tool for resuming S3 uploads as well.