Check whether the source data or the tape drive path is the troublemaker.

The easiest thing to do that is a "bigasm" backup or running a manual backup from the client for a huge file.

For details about "bigasm" please see "Performance Tuning Guide".

the bigasm test also proved to be same from this backup client.

while backups were running on this test_grp one, I copied the backup client 4 savets into another group test_mj_grp and started a backup

now I see that 8 savestreams are sending data @1600KBps,

This points to the fact that the "connection" is bad.

Of course you can also verify that if you copy the files using another method.

If you achieve a similar result, it is obvious that there is something wrong with the network (configuration) - if the hardware itself is fine, it could well be that the traffic is routed in a strange way causing this limitation.

May I suggest you check the fundamental issues first before you rerun a test with NetWorker.

we escalated to DC team, who never told us what they did, but current transfer speed is 35000KBps. thanks for your concern.bingo

You can try to speed up this backup with parallel save stream feature now...