HP NonStop G-Series - Manual (Page 55)

One-time Synchronization
HP AutoSYNC User’s Guide—522580-016
3-17
Compression
exist; the PERSISTENT ON attribute retains the SAFEGUARD file-protection record
after the file is purged.
Compression
To reduce the amount of data that is transferred between the source and a remote
destination, AutoSYNC can compress the data. The LZW compression algorithm
typically eliminates 50 percent to 60 percent of the volume of the data.
The use of data compression requires more cpu time on the source and destination
systems, but reduces usage of the transport medium. By default, AutoSYNC does not
use compression because in most configurations compression does not enhance
throughput. In most environments, the increase in CPU usage outweighs the savings in
data volume.
The use of compression can be set globally through the global parameter
[AUTO]SYNCCOMPRESS or, for a specific synchronization, through the COMPRESS
option.
Process Priority and Throttling
You may specify a process PRIORITY for a synchronization to ensure that
synchronization does not impact higher priority work. Specifying the PRIORITY might
be especially useful if you have enabled compression.
Another way to reduce the impact on other work is to specify a throttling RATE, which
limits AutoSYNC to being active for some percentage (1 percent to 100 percent) of the
elapsed time. AutoSYNC will pause for short periods to reduce its activity, with a
corresponding increase in the time required to complete the synchronization.
Recovery
While a replacement file is being transferred, AutoSYNC retains an existing destination
file by renaming it to a file named AAnnnnnn and placing it in a subvolume with the
MapDB name (normally ASYMAPDB) on the destination file set volume. Partitions,
alternate indexes and protection views are named ABnnnnnn, ACnnnnnn, and so on.
After the synchronization of a file is completed, the renamed recovery file, AAnnnnnn,
is purged unless a failure occurs. If the new file cannot be placed in the destination
location or a failure leaves the replacement file in a corrupt state, the saved version is
used to recover the file.
An archive index file named AAINDEX in the MapDB subvolume contains the mapping
of the original filename to the AAnnnnnn name.
This feature is controlled with the [AUTO]SYNCRECOVER global parameter. By
default, recovery is enabled. You can enable or disable recovery for a specific file set
or for a one-time synchronization by setting the RECOVERY attribute with ADD or
ALTER SYNCFILESET or specifying it as an option in the SYNCHRONIZE command.