Failed to collect disk files location data. Timeout exceeded.
Challenge
A Backup from Storage Snapshot (BfSS) job in Veeam Backup & Replication (VBR) for a vSphere virtual machine (VM) fails with the error:
Failed to collect disk files location data. Timeout exceeded.
Cause
This error is displayed when the disk file location collection task, Map Disk Region, within the vSphere environment takes longer to complete than the timeout for that task allows.
When backing up a vSphere VM using BfSS, VBR has the ESXi host retrieve metadata about the layout of VM disks on the storage (physical addresses of data blocks). The timeout instituted by VBR ensures that if this task takes longer than expected or gets stuck, the individual VM task will fail in order to prevent the job from stalling and delaying the processing of other VMs in the job.
Solution
Investigating Slow Disk File Location Collection
The performance of the disk file location collection is directly tied to the Map Disk Region events that occur within the vSphere environment, both their quantity and how quickly they are completed. Each Map Disk Region task maps the beginning and end point of a contiguous section of blocks on the storage associated with the VMDK. Therefore, if a snapshot is present before the job begins or the VMDK is heavily fragmented, the number of Map Disk Region tasks increases.
The following are likely to increase the number of Map Disk Region tasks that will occur, or how fast they are completed:
- A snapshot exists on the VM prior to the job starting.
Processing a VM with a snapshot present is fully supported; however, that snapshot's presence will impact the "Collecting disk file location" stage of any BfSS job. - Fragmentation of the VMDK on the datastore (not guest level).
- Issues with the storage underlying the datastore.
Adjusting Timeout
If, after investigating the list of potential causes above, the timeout continues to cause the task to fail, you may choose to change the timeout by creating and adjusting the registry value below on the Veeam Backup Server.
Key Location: HKLM\SOFTWARE\Veeam\Veeam Backup and Replication\
Value Name: SanCollectDiskMappingTimeoutSec
Value Type: DWORD (32-Bit) Value
Value Data (Dec.): 10800 [default]
Notes:
- No reboot or service restart is required; the new timeout value will be used during the next job run.
- The value data is in seconds. The default value is 10800, which is equal to 3 hours.
More Information
Tips and Advice
- Considering that a snapshot being present will nearly always drastically increase the number of Map Disk Region tasks that must occur, customers using Backup from Storage Snapshot are strongly encouraged to monitor active snapshots closely.
- As an alternative to creating snapshots on VMs during maintenance, consider using Quick Backup, which can be triggered directly from the vSphere Client using the Veeam Plug-in for VMware vSphere Client.
- When adjusting the timeout value, if you are unsure how high it needs to be set, set it to 86400 (1 day) to allow the task plenty of time to complete. Then, review the job after it is completed and identify how long the "Collecting disk file location" task took, add 10%, and reconfigure the timeout to that discovered value.
To report a typo on this page, highlight the typo with your mouse and press CTRL + Enter.