User's Manual

hp StorageWorks File System Extender User Guide 131
Administration and backup job priority calculation
For administration and backup jobs, the priority is calculated by the following formula:
Migration, recall, recovery, and maintenance job priority calculation
For migration, recall, recovery, and maintenance jobs, the priority is calculated by the following formula:
NOTE: You can manually increase or decrease a job priority using the fsejob --priority
command.
Example of resource allocation
The following is an example output of the fsejob --list command, which shows the resource
allocation queue with job priorities. Note that jobs that are currently not waiting for resources do not have
priorities assigned.
Multiple copying
FSE migrates data to FSE media to maintain an offline copy of the data on HSM file systems. Each file on
an HSM file system has a corresponding copy on an FSE medium. The medium that stores such a copy
belongs to the FSE media pool assigned to the respective FSE partition. Therefore all files on a particular
HSM file system are migrated to the same media pool. In case of a disaster, this has the following
advantages:
• A complete HSM file system can be restored from the FSE media, including the file system structure for
the migrated file data.
• The HSM file system restore period is reduced since only the active (online, frequently used) file data
need to be restored.
In order to enhance security and eliminate the risk of physical damage to your media, which can result in
data loss, you can configure FSE to make redundant copies of the file data during migration jobs. To do
this, you need to configure additional media pools, allocate additional media, and assign them to the
particular HSM file system. All media pools must be configured with the same block size (BlockSize
variable in the FSE media pool configuration file).
Priority = JobTypePriority
+ (ElapsedTime * TimeStep * JobTypePriority/10000)
+ (Phase * PhaseFactor)
Priority = PartitionPriority * JobTypePriority
+ (ElapsedTime * TimeStep * JobTypePriority/100)
+ (Phase * PhaseFactor)
cumulus:/root # /opt/fse/sbin/fsejob -l
JobId Part JobType Prio Started Status
20051119000043 Part01 Migration n/a 2005/11/19 08:02:05 Writing to Medium
20051119000044 Part01 Migration 1023 2005/11/19 08:02:40 Waiting for resources
20051119000045 Part01 Migration 1009 2005/11/19 08:03:15 Waiting for resources
20051119000046 Part02 Recall 4000 2005/11/19 08:03:38 Waiting for resources
20051119000047 Part01 Migration 1030 2005/11/19 08:02:40 Waiting for resources
20051119000048 Part02 Migration 1016 2005/11/19 08:03:15 Waiting for resources
20051119000049 Part01 Recall n/a 2005/11/19 08:03:38 Reading from Medium
20051119000050 Part01 Migration n/a 2005/11/19 08:04:30 Writing Disk Buffer