Ron, Ron and Jenny Hawkins wrote:
<snip>
I have several. My favourite is handling video for a casino. There are 800+ cameras feeding 200 DVR. Every 15 minutes the DVR will dump their contents to disk, and those disk files must be backed up to tape before the 15 minutes is over. This was two years ago, but if throughput fell below 1.2GB/sec the whole process would fall apart. I just had a new request last week for backup to run at 1.6GB/sec without affecting OLTP at 25% of peak OLTP load (25000 IOPS) while doing two way remote copy in a 30% cache hit workload. And this is in a 3rd world country!
You are getting 1.2GB/sec on a single tape? What type of tape devices were used? Or is this a aggregate based on 200 tapes at 60MB/sec each?
<snip>
Ask Felix Unger about assume!
Yes I know about assume and unfortunately I keep doing it. :)
<snip>
You're asking the wrong person here. I am a confessed TMM bigot, and I fail to see why Mainframe sites would go VTS when TMM is so much faster, simpler and easier.
I guess that is why we are a TMM shop and not a VTS shop also. No matter what IBM tried to do, it always ended up costing us more money for a VTS system than what we were spending now.
Ron
John. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
