Re: Why the All-Time & 1998 Top-Ten $tpmCs use TUXEDO?

Paul, Allow me to respond to your points, as they are all good ones. Paul Ilechko wrote: > > The answer to the question in the header is three-fold. > > 1. Tux has special pricing for a cut down verision for benchmarking, > which keeps cost per TPCC down > It is absolutely false that the vendors are using a version of TUXEDO that is "Cut Down" in any form whatsoever. This is the same software that every TUXEDO customer runs. That's a pretty nasty accusation, and it would violate the TPC rules. Curiously, the price for TUXEDO in the benchmark results is significantly higher than the price that your company, IBM, lists for Encina (TXSeries), so I'm not sure what your complaint is. > 2. TPCC is typically run on a single machine, so doesn't hit all > Tuxedo's scalability problems as Domains increase. > I don't know what scalability problems you are referring to. TUXEDO Release 6.4 made significant improvements to the Domains performance. If you have specific examples, please post them, or have the customer contact our Customer Support line so that we can help them resolve these problems. Otherwise, your statement lacks credibility. You can read numerous customer success stories at http://www.beasys.com/action/css/index.htm These are real applications running under real-world conditions. > 3. Tux is relatively easy for the initial install, so it's appealing > to the hardware vendors and DB vendors who run most of the benchmarks. Thank you. > Of course, this has nothing to do with the real world issues of > reconfiguring your /WS clients each time you add new gateway servers. If you hard-code the network address, then maybe your client has to change. Same for any Network application. But if you retrieve the addresses dynamically from a directory, no change should be needed when a gateway address changes. > > And of course, this is also incorrect, as THE top TPCC benchmark uses > TXSeries and Oracle on an SP2. > I believe that you haven't checked for a while, since the SP2 result is only #3 in performance. The #2 result uses TUXEDO on a Sequent system. You can see the Top 10 table at http://www.tpc.org/new_result/ttperf.idc However, the topic of this posting is Price/Performance (note the '$'), and it is definitely correct. I recommend that you speak to your company's TPC representatives before making statements such as this in a public forum. This way, you can verify your information in advance. > Paul. > > *************** > > Paul Ilechko > http://www.transarc.com/~pilechko/homepage.htm > Scott Orshan BEA Systems TPC Representative 908-580-3185 sdo@beasys.com