Comprehensive data protection for all workloads
Post Reply
Woffle99
Novice
Posts: 8
Liked: never
Joined: Feb 21, 2014 4:44 pm
Contact:

Large bandwidth use from backup copy target back to Server

Post by Woffle99 »

Hi,

I've recently noticed that we have extremely high bandwidth usage on our broadband connection, but it is FROM the remote location that stores our copy backups back to the Veeam server. I have a job that runs every week to copy the latest backups to the remote location which runs fine and which is configured so that it does not run during regular working hours. However, I can see from figures from our ISP that there is a massive data transfer the other way which seems to happen every four weeks and is not restricted by which time it runs. It saturates our connection and massively slows down all other traffic across our WAN. Judging by the size of the data transferred, it would seem to me that it is transferring the entire remote storage location back to the main Veeam server for some reason, presumably for some kind of check or verification.

How do I get it to stop doing this? I have tried the Global Throttling rules and it ignores them completely, as it ignores the scheduling on the job. I have turned off the health check setting that was turned on (and was set to every four weeks, but then was set to a Sunday and the bandwidth drain only ever starts on every fourth Tuesday) and restarted the job, but it is now saying that it is going through 'backup file verification' and it is still killing our connection.

This is ridiculous - what is the point of having settings that restrict the times and bandwidth but which only appear to take effect when copying data TO the remote location, not pulling it FROM it (for whatever reason)

I am on Veeam 7.0 - please can anyone help. I have opened support case 00817130 but am waiting for a reply so thought I would also post here.

Thanks,
Mark
foggy
Veeam Software
Posts: 21069
Liked: 2115 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by foggy »

Mark, what kind of target repository do you have? Is it CIFS share by any chance? In this case, if there's no gateway (proxying) server specified for this repository, verification traffic will go through the Veeam B&R server. However the fact it ignores scheduling looks unexpected and this is what I would recommend to investigate via support.
Woffle99 wrote:This is ridiculous - what is the point of having settings that restrict the times and bandwidth but which only appear to take effect when copying data TO the remote location, not pulling it FROM it (for whatever reason)
And this also makes me think that there's no proxying server in the target location, since throttling rules (which are actually two-way) are effective between two Veeam data mover agents, while in case of CIFS repository without the proxying server the target data mover agent is started on the backup server itself.
Woffle99
Novice
Posts: 8
Liked: never
Joined: Feb 21, 2014 4:44 pm
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by Woffle99 »

Hi Foggy,

Yes it is a CIFS share and no, there is no server in that location so the verification would presumably need to be done on the Veeam server at the source location. That would be fine if, as you say, it paid attention to either the network throttling or the job scheduling settings, but that only seems to apply to traffic going from source to target, not vice-versa.

I've tried turning off the health check but when it re-submitted the job, it still said it was 'Starting backup file verification' and the stream of data from target to source carried on.

Cheers,
Mark
foggy
Veeam Software
Posts: 21069
Liked: 2115 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by foggy »

Add some Windows server at the target site as a gateway for the repository and see whether this helps.
Woffle99
Novice
Posts: 8
Liked: never
Joined: Feb 21, 2014 4:44 pm
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by Woffle99 »

Unfortunately there is no windows server at the target site - the only devices there is a netapp NAS which is what the CIFS share is located on and a couple of desktops/laptops.
foggy
Veeam Software
Posts: 21069
Liked: 2115 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by foggy »

It can be a desktop OS also. Basically, for optimal backup performance and to avoid unnecessary back and forth data flow over slow link, it is always recommended to have a proxying server located close to remote CIFS repository.
Woffle99
Novice
Posts: 8
Liked: never
Joined: Feb 21, 2014 4:44 pm
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by Woffle99 »

Thanks Foggy. I'll look into that, but it might be a while before I can get a desktop configured and installed up there (it's quite a distance away from me). If this is verification traffic, is there no way of turning it off? I would have thought dropping the healthcheck would have done something, but it doesn't seem to have made a difference.
foggy
Veeam Software
Posts: 21069
Liked: 2115 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by foggy »

Disabling health check should work, this is also a subject for investigation with support.
Woffle99
Novice
Posts: 8
Liked: never
Joined: Feb 21, 2014 4:44 pm
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by Woffle99 »

I disabled the job, re-edited the settings and then re-booted the server before re-enabling the job and it seems to have done the trick. This time it did not start with the 'backup file verification' but immediately paused waiting for the scheduling window (overnight).

I do feel nervous about not having the healthcheck though and I do feel that the check should obey, at the very least, the same scheduling restrictions as the job proper (i.e. only running overnight and weekends in my case) even if it is not possible to limit the bandwidth without a proxy at each end. Something for a future update hopefully?
foggy
Veeam Software
Posts: 21069
Liked: 2115 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: Large bandwidth use from backup copy target back to Serv

Post by foggy »

Actually we've already checked this on v8 and everything looks fine with this version in terms of complying with the specified window. So probably some issues in v7.
Post Reply

Who is online

Users browsing this forum: No registered users and 187 guests