Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations Rhinorhino on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
big volumes
- Thread starter dmpe
- Start date
- Thread starter
- #4
I have 2Tb IDE -> SCSI RAID connected to solaris/sparc8
with legato network edition 6.1
Initially I want create many volumes with size ~20-30Gb (like
DD3 tapes which I used before). But (thanks 605) then I find out
that my network edition support only 16 devices. So I want create
16 big volumes.
I want do 4 groups with from 2 to 6 devices in each. Some groups
will be backup >10GB each day with thousands of files..
with legato network edition 6.1
Initially I want create many volumes with size ~20-30Gb (like
DD3 tapes which I used before). But (thanks 605) then I find out
that my network edition support only 16 devices. So I want create
16 big volumes.
I want do 4 groups with from 2 to 6 devices in each. Some groups
will be backup >10GB each day with thousands of files..
O.K. - you really want to backup to big (advanced) file devices.
In general there is no technical problem. NetWorker can do this. In fact, you must split
your source data to create multiple save streams which then can be saved in parallel.
However, your environment must be very carefully designed - if not, you will not be able to
increase the throughput. I think this is what you are really looking for.
For example, if only one stream can feed the backup device so fast that it runs with maximum
speed, there is no increase for the throughput if you work with multiple save streams.
For tape drives, the limit can be determined easily, for disk devices, this is much more difficult.
Most likely not the devices but the data paths will become the bottleneck. To find out the
optimal scenario, most likely only practical benchmark tests can help.
In general there is no technical problem. NetWorker can do this. In fact, you must split
your source data to create multiple save streams which then can be saved in parallel.
However, your environment must be very carefully designed - if not, you will not be able to
increase the throughput. I think this is what you are really looking for.
For example, if only one stream can feed the backup device so fast that it runs with maximum
speed, there is no increase for the throughput if you work with multiple save streams.
For tape drives, the limit can be determined easily, for disk devices, this is much more difficult.
Most likely not the devices but the data paths will become the bottleneck. To find out the
optimal scenario, most likely only practical benchmark tests can help.
- Status
Similar threads
