Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations strongm on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Terradata Platform
- Thread starter JimOwens
- Start date
connectxyzzy
Programmer
To my knowledge, Teradata is a storage subsystem (usually high-end and expensive). If they are into other hardware such as servers, I am not sure.
I <i>believe</i> a terradata platform is merely a name for a server/computer with >1024GB of starage space; but don't quote me on that. 
<p>-Robherc<br><a href=mailto:robherc@netzero.net>robherc@netzero.net</a><br><a href= > </a><br>*nix installation & program collector/reseller. Contact me if you think you've got one that I don't
<p>-Robherc<br><a href=mailto:robherc@netzero.net>robherc@netzero.net</a><br><a href= > </a><br>*nix installation & program collector/reseller. Contact me if you think you've got one that I don't TerraData is a High Availability Client/Server storage system, basically it has the capability for "dual" everything and redundant everything, including mirroring. The hardware in the past has not always been the most reliable but with the efficient redunancy features this is not neccessarily an issue.
All of the above are incomplete at best, the first two are incorrect. Teradata is a massively parallel relational DBS. It uses standard sql but has enhanced sql, as do other vendors such as IBM's DB2/UDB.
It currently runs under MP-RAS UNIX, spawning processes that perform as processors. V2 is the current O/S and DBMS. Under V1, 1985-1994 approximate, Teradata from Teradata Corp. was made up of proprietary hdw, o/s and dbms, and a proprietary high speed bus called the YNET. NCR purchased Teradata Corp in the mid-90s; most of the Fortune 500 and many other companies quietly use it. Minimal marketing occurs; that seems to be why MIS types don't know about it.
It is called massively parallel because it will take a request, and send it to multiple processors who each call a subset of the answer set from its own disk drives. The operation is performed - read/sort, updates, deletes, inserts - and the answer sets if any are returned to the sender, and all changes to the rows are applied to each processor's disk drives. It can support scores of terabytes because it is modular, you add unix/processor nodes and disk drives as needed!
It is true that it is highly fault resilient, beyond fault tolerant, in its current incarnation as V2 [version 2]. It can, if installed and planned correctly, be totally fault tolerant, with loss of data availability only occurring if Raid 5 is used w/o Fallback. Fallback is a complicated data protection scheme, but easily implemented and transparent to the programmer. Fallback writes 2 rows to 2 different disks for each data row. My mission critical system uses Fallback on top of Raid 5.
It is a very sophisticated Back-end RDBMS, relieving the mainframe hosts from processing cycles. It can be connected CONCURRENTLY to multiple networks, mult. hdw clients, several MVS images and even multiple MVS complexes.
It's slick, and has been an industry leader since its inception, because it is continually improved.
The next generation runs under NT [the dbs runs even on a desktop] on any server hdw that will run NT.
Contact me for more info, or send me your phone number.
It currently runs under MP-RAS UNIX, spawning processes that perform as processors. V2 is the current O/S and DBMS. Under V1, 1985-1994 approximate, Teradata from Teradata Corp. was made up of proprietary hdw, o/s and dbms, and a proprietary high speed bus called the YNET. NCR purchased Teradata Corp in the mid-90s; most of the Fortune 500 and many other companies quietly use it. Minimal marketing occurs; that seems to be why MIS types don't know about it.
It is called massively parallel because it will take a request, and send it to multiple processors who each call a subset of the answer set from its own disk drives. The operation is performed - read/sort, updates, deletes, inserts - and the answer sets if any are returned to the sender, and all changes to the rows are applied to each processor's disk drives. It can support scores of terabytes because it is modular, you add unix/processor nodes and disk drives as needed!
It is true that it is highly fault resilient, beyond fault tolerant, in its current incarnation as V2 [version 2]. It can, if installed and planned correctly, be totally fault tolerant, with loss of data availability only occurring if Raid 5 is used w/o Fallback. Fallback is a complicated data protection scheme, but easily implemented and transparent to the programmer. Fallback writes 2 rows to 2 different disks for each data row. My mission critical system uses Fallback on top of Raid 5.
It is a very sophisticated Back-end RDBMS, relieving the mainframe hosts from processing cycles. It can be connected CONCURRENTLY to multiple networks, mult. hdw clients, several MVS images and even multiple MVS complexes.
It's slick, and has been an industry leader since its inception, because it is continually improved.
The next generation runs under NT [the dbs runs even on a desktop] on any server hdw that will run NT.
Contact me for more info, or send me your phone number.
- Status
Similar threads
- Locked
- Question
- Locked
- Question