At one of my customers they experienced some issues with Veeam Backup and Replication v7. During the search for what was the problem causing this behaviour and the way to solve it, I noticed their Veeam environment wasn’t updated to the latest Patch level, which is Patch 4.
With Patch 4 Veeam solved multiple problems, and it solved the issues the customer had. So before you troubleshoot like hell, make sure the Veeam Backup and Replication v7 Patch 4 is installed.
Some of the issues solved include:
- Under certain circumstances, the target data mover may crash with the “Cannot allocate memory for an array” error while processing a backup job with WAN storage optimization setting.
- Increased maximum amount of open files for Linux data mover to improve backup reliability in cases with a very large amount of concurrent tasks hitting the same repository.
- If the same ESX(i) host is added to Managed Servers both as a part of the vCenter, and as a standalone host, the jobs fail to remove VM snapshots correctly.
- Replication job does not truncate Microsoft Exchange transaction logs when the corresponding setting is set to “Truncate logs on successful backup only“.
- Setting EnableSameHostHotaddMode registry value (introduced in Patch #3) to 1 causes replication jobs to fail with the “Object reference not set to an instance of an object” error.
- If you clone replica VM with vSphere, and start the replication job, the target VM will be selected randomly between the two VMs.
And some of the new enhancements and features:
VMware Virtual SAN (VSAN)
- In addition to adding basic support (as provided by other vendors), the intelligent load-balancing engine was enhanced to account for VSAN specifics. As the result, for each VM the job will pick backup proxy running on VSAN cluster node with most of the virtual disks’ data available locally. This significantly reduces backup traffic on VSAN cluster network, resulting in minimum possible impact on production environment from backup activities.
- The maximum allowed amount of restore points in the Backup Copy job has been increased to 999.
- Backup Copy will now resume the transfer after network connection drop to a Linux-based backup repository.
- Backup Copy jobs should no longer report errors during the days when source backup jobs are not scheduled to run – for example, during the weekend.
More information and download can be found here