Backups fail with timeout error

[MartijnV]

My company is testing with Veeam backup at the moment, but we are running into some problems.

I installed Veeam backup and restore 6 on a VM with 8 cores and 8gb memory. I attached a 5tb iSCSI target to this server as drive K: that I want to use to for storing backups.
On our first ESX server we have about 25 VM's, I created jobs for every seperate VM and set all jobs to run daily at 10pm.
Backup proxy is set to 4 concurrent tasks, Transport mode and connected datastores on auto, Backup repository is set to "Microsoft Windows Server" (pointing to the drive K:), limit the maximum concurrent jobs is set to 4 also.
All hardware is connected with 1Gb network.

When I manually start the jobs 1 by 1 everything seems to work just fine, but when I start them all at once some jobs will time out before starting.
I found a Kb article on the veeam site saying that I should just reschedule the jobs to not start at the same time.
Should I really schedule all tasks at different times? Or should the amount of concurrent tasks set on the backup proxy (and/or on the repository) take care of this?

[Gmc85]

Put more of the VM's in the same jobs (eg 5 jobs with 5 vm's each) and stagger them slightly - there's not much need to have a different job for each VM!

[Vitaliy S.]

The amount of concurrent tasks is in fact controlled by the backup proxy/repository servers, however if the jobs that were put on hold failed to start processing VMs with a defined period of time they will timeout and the default timeout in 6.1 is 23 hours.

[MartijnV]

I will try again with more VM's per job.

@Vitaly: Is there a way to change the timeout values? Backup reports show timeouts after 360ish seconds.

Btw is it recommended to have the SQL database on a seperate server?

[Gostev]

I recommended to have separate SQL database when you have all-in-one install (all components on the same computer), as backup proxy server is very CPU hungry and may potentially affect other applications. Actually, version 6.1 reduces impact of backup proxy activities on other applications, so you may as well want to upgrade to 6.1 as a first step. Thanks!

[Vitaliy S.]

I'm not sure why your jobs behave like that, as I have just verified with the QC team that we do not have a 360 seconds timeout for the backup/replication jobs.