I cannot believe you could make such an uninformed comment - DO you really think that large data entres have a bunch of clustered Windows Servers? SNA/Token Ring is still the preferred method for large> data centre's topology because its secure. Not in this country. Its TCP/IP directly to the
On Friday 15 June 2007, Registration Account wrote: mainframe. Security by obscurity has been entirely discredited here. We don't rely on a silly protocol change to provide security. If that's the best you got, I'm selling my ANZ stock first thing monday morning.
John the last thing I would do is devoid yourself of ANZ stock. I have been in ANZ data centre and its NOT TCP/IP directly into the mainframe. I can promise you
Registration Account wrote: that ANZ is a very very very secure installation and works much the same as the below information. The people at the branches sit in front of a 3270 or derivation emulation screen (SNA) and all internal links are via an SNA server at each branch. We do it the right way. There are a few exceptions where some overseas terminals and some branch PC run a TCP/IP GUI - Windows 200 I believe -but they are secure as the data stream in ALL encrypted. ANZ Bank Data Centre is located in Melbourne, Australia and I think you are in New Hampshire US or you ISP - Dynamic Network Services, Inc. - is located there. Don't ever worry about any institution where the users predominately site behind a 3270 or derivation emulation screen as this means they are using a Mainframe with secure SNA comms. The term silly protocol change is not valid, its hard, dam hard and that why we need another mainframe to not only convert the protocol but to make the transfer transparent and remove TCP's ability to escalate authority. This aspect is O/S dependant and the Mainframe that performs the change runs VTAM and its as safe as a house. John, yes there is online banking as we both know, However there is a dirty great big IBM 3172 sitting between all main internal SNA architecture that is used by every branch ANY TCP source must pass through the 3172 is to convert its native SNA at the Data Centre. This provides connectivity between the protocols. The 3172 runs a MVS/C interface initially, but has been replaced by VTAM. The 3172's many of which are still operational have been replaced by new Models as support from IBM halted in 1988. They still remain the workhorse for transparent TCP/IP/SNA protocol. I cannot believe you could make such an uninformed comment - DO you really think that large data centres have a bunch of clustered Windows Servers? SNA/Token Ring is still the preferred method for large data centre's topology because its secure. I have worked with some of these data centres around the work and the biggest non-Military Data Centre on the planet before I retired. There are 4 of the biggest Data Centre in the world located on the Planet. 1 in Tulsa Oklahoma - USA, 1 In Denver Chicago - USA, 1 in Swindon UK and another in Asia - location classified. Private satellite links provide the comms which is achieved by SNA/T Each is capable of quad instantaneous redundancy should the second Mirror Mainframe site of any one installation go down. With an ability to process upwards of 1500 transaction per second running an O/S known as TPF/UG on top of assembler or C++ code. Their limited public access is provided by TCP/IP via a translation Mainframe similar to a 3172. John - any serious Data Centre does not use Blade servers or a whole bunch of high powered PC's and entertain this notion is folly and probably based on youthful enthusiasm. (I do not mean this to be a derogatory statement) Have a great day :-) Scott