[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference ulysse::rdb_vms_competition

Title:DEC Rdb against the World
Moderator:HERON::GODFRIND
Created:Fri Jun 12 1987
Last Modified:Thu Feb 23 1995
Last Successful Update:Fri Jun 06 1997
Number of topics:1348
Total number of notes:5438

1327.0. "Benchmarks ?" by MXOV09::EPELAEZ () Fri Dec 17 1993 02:33

	I need information about benchmaks of VMS/RDB against ORACLE, SYBASE, 
ADABAS, (VMS version 5.5x, Hardware don't care). Does anybody know about any 
source with this information (vtx or notes) ?

						Thanks, 
						Enrique. 
T.RTitleUserPersonal
Name
DateLines
1327.1What is the customer's interest?BOUVS::OAKEYAssume is *my* favorite acronymFri Dec 17 1993 17:1520
�                     <<< Note 1327.0 by MXOV09::EPELAEZ >>>
�                               -< Benchmarks ? >-

�	I need information about benchmaks of VMS/RDB against ORACLE, SYBASE, 
�ADABAS, (VMS version 5.5x, Hardware don't care). Does anybody know about any 
�source with this information (vtx or notes) ?

Enrique,

Are you looking for any specific benchmarks?  There is TPC-A (a 
debit/credit type benchmark which DECRdb and Sybase publish), TPC-B (a 
modified debit-credit type benchmark which Oracle publishes), and TPC-C (a 
simple order-entry benchmark which none of the above has published yet 
since it is fairly new).

It would help to know more specifically what you're looking for... 
Comparing databases based on a version of VMS in my opinion probably isn't 
real useful.  Comparing on like hardware platforms and a platform of 
interest to the customer is probably more useful.

1327.2Benchmark informationMXOV08::EPELAEZFri Dec 17 1993 23:3535

	OK, thanks by answer to me. The system in which will be installed the
database system is an VMS CLUSTER with two nodes : a VAX 9000 system and a
VAX-6000, in addition they have in a network a VAX-3400 and a VAX-3800 that  
probably will be used as a database servers for distributed aplications. Then
is important for my customer knows characteristics of  RDB, ORACLE, SYBASE and
ADABAS in this platforms (VAX-9000, VAX-6000, VAX-3400 and VAX-3800). They 
already have developed any applications using ADABAS (with NATURAL), but they 
aren't satisfieds with the respond time of ADABAS, they thing evaluate RDB, 
ORACLE and SYBASE migrating information of his own tables and taking the 
follows times :

	- Loading a table (with aprox. 1000000 rows of 200 bytes each row)
	- Updating all records in same table.
	- Executing any selects instructions of SQL, that represent very 
          frecuent ways to access information for obtain reports.

	I thing this 3 tests is not enough to take a determination and I want
an advise of how (or where) to obtain information to advise to my client about 
a valid methodology of evaluation. 

 	In other aspect they have already information about develop tools as 
ACMS, RALLY for VMS, RALLY for DOS (probably Client/ Server applications be 
implemented), and also about other client tools as Q+E and EXCEL 
(It can access RDB), etc. But a don't know if we have this kind of information 
about our competition (especially ORACLE and SYBASE) and if exist in any 
place (vtx or notes) any comparasion table of this kind of tools.Finally 
I want know if this DataBase systems has equivalents to RdbExpert, 
Graphical Schema Editor and DECTrace.

    					
				Thanks, 
                                (and excuseme because my english is not well)                               
				Enrique.
1327.3Just Do It ?MSDOA::SECRISTPress 6 if you hate voice mail.Tue Dec 21 1993 11:2311
	Have you considered asking for a short-term loan-of-product
	for each of the candidates to perform these operations ?
	While this takes the most effort, it assures the highest
	degree of satisfaction -- and when you're all done you've
	got most of the work done to get the winning database online.
	Benchmarks could be misleading for you (especially from Oracle).

	Regards,
	rcs

1327.4Lies, Damn Lies and BenchmarksBIGRED::SPARKSI have just what you needWed Jan 05 1994 21:4513
    I agree with -.1, however unless they have applications that do lots of
    loads and mass updates the benchmarks you mentioned will also be
    misleading in real world use.  ADAbase in particutular likes to point
    to these particual items because that is where they perform best, Loads
    and mass updates, neither which take into account things like row level
    locking and query optimizer.
    
    Published benchmarks can also cause you lots of trouble unless reported
    by and independent group, which also have their problems because of
    limited knowledge of each product.  Hopefully the TP/C will help sort
    some of this out.
    
    Sparky