|IDMS/SQL News 1.2
Vol 1.2 Technical Information for IDMSâ Users May 1992
Optimizer Technology of CA-DB/SQL scores very high points in an Industry Analysis. 42 features were considered by a study carried out by InfoDB Journal . CA-IDMS/SQL Optimizer is not only based on CA/DB Optimizer but also incorporates existing CA-IDM S technology to ensure OLTP performance on 370 platforms.
Relational DBMS Optimizers
- Based on InfoDB Magazine
|In all, 9 databases - CA-DB, DB2,
Informix, Ingres, Oracle, RDB, Sharebase, Sybase and
Tandem - were evaluated against 42 features. CA-DB scored
25 Yes, 9 Partial and 8 No.
Ingres scored very well too. DB2 scored 16 Y, 9 P and 17 N. All others were far behind.The results are not surprising at all.
CA-DB had its origin in 'Mars' an
original relational DBMS by ESVEL Inc. the company
founded by alumni of System-R prototype (Gartner Group
report SMS R-106-105 - Feb 14, 1990). IDMS-DB mainframe
optimizer is an
System-ID in Sysgen
In the IDMS/DC Sysgen, there is a new parameter coded
as 'system id is
Transaction Statistics more extensive in 12.0
DBNAME, SEGMENT and
DBTABLE consists of DBNAMES and is defined in OCF. Each DBNAME consists of one or more database SEGMENTS. Segment in turn contains one or more physical files/areas.
DBNAME was an 'optional' feature in 10.2, used only if the user had multiple databases or secondary dictionaries. In 12.0, DBNAME is always required and is closely associated with database SEGMENTs. Then there is the question of all 10.2 applications where DBNAMEs were not used. For this reason, 10.2 compatibility is provided at the top of DBTABLE by supporting subschema mapping. This will be helpful in the sense that no application program re-writing is required if no dbnames were used in 10.2.
(refer to DBTABLE display box below)
The first subschema mapping with IDMSNWK? is a special case which achieves two things in one shot. First it establishes the default dictionary as SYSDICT. ie if no dictname is given or DICTNAME ='', system will always use SYSDICT. Note that in R12, sys gen dictionary is always SYSTEM, which as you see is NOT the default dictionary.
Secondly, if you try to read the dictionary database, using IDMSNWK? subschema, then database defaults to DBNAME= SYSDICT. ie we always read the SYSDICT DML area. In the same way, for the user database EMPSCHM, if no dbname is specified, subschema EMPSS01 will always default to DBNAME= EMPDEMO. This is equivalent to setting 'DCUF SET DBNAME=EMPDEMO' or 'MOVE 'EMPDEMO' to DB-NAME' in ADS. The following discussion is based on 12.0 OLQ access against 1) employee database using EMPSS01 2) dictionary database using IDMSNWKA
DICTNAME ' ' DBNAME ' ' (or nothing is set)
DEPARTMENT REPORT 05/05/92 DEPT-ID-0410 DEPT-NAME-0410 5300 BLUE SKIES 5100 BRAINSTORMING 2000 ACCOUNTING AND PAYROLL 1000 PERSONNEL 3100 INTERNAL SOFTWARE 3200 COMPUTER OPERATIONS 4000 PUBLIC RELATIONS 1000 EXECUTIVE ADMINISTRATION 5200 THERMOREGULATION END OF REPORT
The sample DBTABLE DISPLAY in Release 12.0: DBNAME *DEFAULT MATCH ON SUBSCHEMA IS OPTIONAL SUBSCHEMA IDMSNWK? MAPS TO IDMSNWK? USING DBNAME SYSDICT SUBSCHEMA IDMSCAT? MAPS TO IDMSCAT? USING DBNAME SYSDICT SUBSCHEMA EMPSS01 MAPS TO EMPSS01 USING DBNAME EMPDEMO DBNAME SYSDICT MATCH ON SUBSCHEMA IS OPTIONAL SEGMENT APPLDICT SEGMENT SYSMSG SEGMENT SYSSQL DBNAME SYSDIRL MATCH ON SUBSCHEMA IS OPTIONAL SEGMENT SYSDIRL SEGMENT SYSMSG DBNAME SYSTEM MATCH ON SUBSCHEMA IS OPTIONAL SEGMENT CATSYS SEGMENT SYSMSG SEGMENT SYSTEM DBNAME EMPDEMO MATCH ON SUBSCHEMA IS OPTIONAL SEGMENT EMPDEMO (DCMT D DBN display
IDD is enhanced to include SQL Catalog
Integrated Data Dictionary is enhanced to include SQL
Catalog and Security Catalog. It is fair to say:
SQL Benefit for Network Databases
Release 12.0 support for SQL access against existing
network databases provide some unexpected side benefits.
One of them is the ability to access multiple network
databases in the same program - ADS, COBOL or PL1.
SELECT * from EMP.EMPLOYEE E, POST.POST_OFFICE P where EMP_ID = 1023 and E.POSTNR = P.POSTNR;
In OLQ 10.2, this was possible after signing on to the
two subschemas. Major difference in R12.0 is that you can
do this in a program, online and batch, not just in query.
Also it can be incorporated into a DECLARE CURSOR SQL
statement, for later processing using FETCH.
Self-Training on SQL
A sample SQL database with 12 tables is installed as
part of R12 base tape. One can use SQL Self Training
Guide to test and learn SQL statements against this
CREATE SCHEMA CJD DEFAULT AREA TEST_AREA; CREATE TABLE CJD.S ... CREATE TABLE CJD.P ... etc. INSERT INTO CJD.S VALUES ('S1','Parts','London');
Another person can do:
CREATE SCHEMA DBXY DEFAULT AREA TEST_AREA; SET SESSION QUALIFIER DBXY; CREATE TABLE DEPT..... CREATE TABLE EMPL... or in IDD or OCF ADD MODULE CJD_TABLES LANGUAGE IS OCF MOD SOURCE FOLLOWS CREATE SCHEMA ... CREATE TABLE ... CREATE INDEX... ETC. ETC. MSEND;
Basically an AREA (=tablespace) is shared by all the schemas. In the initial stages, this will be extremely useful for individual programmers who want to try their SQL expertise. DBA need not create individual AREAS or FILES etc. And the programmer s need not spend time waiting for DMCL definitions and file allocation.
European IDMS User Association Executive meeting was held in Helsinki on 24/04/92. CA was represented by Lars Ahlberg and Scott Corrigan, CA France (IDMS European Product Champion) among others. Wishlist for Europe, 91 was presented and out of 105 r equests from IDMS Customers 70% were favourably considered by CA. They are either already implemented in the new release or accepted for a future release.
Release 12.0 GA tape will be available in September 92 timeframe. Sincework on IDMS 12.1 is already on the way, some technicians had mistakenly mixed up 12.1 beta tape with the GA availability of R12.0. In fact, some customers have plans to go into production with ESP tape in June-July period and these operations will allow CA to monitor IDMS in a production environment.
IDMS User Association World Meeting was held in Dallas, April 5-9. The event was a great success with real life presentations from users, product demos and future directions from CA, speaches by industry gurus and analysts.
Information Management Conference (IMC). Users can choose among 3 groupings- IDMS, Tools, and Datacom .
Back to the Main Page
This page at http://www.reocities.com/idmssql/idms12.htm
IDMS/SQL is published on behalf of IDMS WatchDog Group, -Helsinki-Oslo for free circulation among IDMS Users Worldwide. IDMS/SQL News is not a CA publication. CA-IDMS/DB, CA-IDMS/DC and CA-ADS are registered trademarks of Computer Associates International Inc. CICS, IMS-DB/DC and DB2 are registered trademarks of IBM Corporation. Technical examples are only guidelines and modification might be required in certain situations and operating systems. Opinion expressed are those of the authors and do not represent the views of IDMS clients or related vendors. © Permission is hereby granted to use or reproduce the information, only to IDMS Customers and Consultants, provided the material is distributed free of charge.