I am getting vastly overstated speeds running the sample script q_ses_stats
on TSM 6.1.3.1 - works fine on v5. Example -  a virtual volume session on v5
(the sending session) gives me about 51 MB/sec which sounds very plausible.
The receiving node on v6 gives me about 480 MB/sec which looks like it is
out by a factor of 10 - sometimes it is overstated even higher. The data
flow is from an LTO4 tape via 1GbE to LTO4 tape (ba stg backuptape copypool
where copypool is based on virtual volumes at the source server; data is
sent directly to tape at the target server; steady state where 100s of GB
are being backed up).  I did a select * from sessions on v6 and it looks OK.
Seems like somewhere in the arithmetic something goes wrong. Has anyone else
experienced this? If so is there a fix?



Joerg Pohlmann

250-585-3711