NetBackup™ Backup Planning and Performance Tuning Guide

Accelerator for file-based backups

NetBackup Accelerator worsk with file and folder backup policy types Standard, Windows, and NDMP.

NetBackup Accelerator uses files called track logs to enable tracking changes at a file and sub-file (segment) level on a policy/client basis. The track logs may be stored in different locations based on policy type and how a backup is performed. Individual track logs are created for each policy/client combination.

The track log size is relative to the size of the data (number and size of files) being backed up. So, enabling accelerator for backups of very large file servers or NAS may result in needing to manage the storage space for track logs.

The file system location where the track logs are stored needs sufficient space and performance characteristics for track log use. The track log is updated when an Accelerator backup is running. If the storage where the track log is located has high latency (read or write) or heavy I/O load, the performance of the backup may be negatively affected. Also, if multiple backups are running concurrently on the system (client or media server) where the track log is located that could be a source of I/O contention and should be monitored if there are performance concerns with Accelerator backups. Insufficient space for track logs may result in backup failures or the Accelerator optimization being disabled for the backup.

The following formula is used to compute track log size:

Track log size in bytes = (number of files * 128) + (((total disk space used) / 128KiB) * 20)

For example, if you have 179,418,964 files and 28,264,740,674 bytes in total disk space used, the result is:

(179418964*128) + (((28264740674/128)*20)) = 22965627392 + 4416365730.3125 = 27381993122.3125 bytes

More information is available about track log size:

See Controlling disk space for Accelerator track logs.