DB2 LUW 11.1 product documentation available in PDF format

For a period of time the DB2 LUW 11.1 documentation was only available online. The good news is it's now available in PDF format and can be downloaded to read through . To download the documentation visit the Db2 Version 11.1 for Linux, UNIX, and Windows English Manuals Read more →

Troubleshoot Long running sql statements with LONG_RUNNING_SQL view

A customer called me yesterday and complained about slow response time on a DB2 database. I asked a few questions and they mentioned a few ad-hoc queries were executing. If I’m doing production troubleshooting – running a sql trace is a powerful method to retrieve detailed information. Prior to getting the detailed information , I’ll use the SYSIBMADM.LONG_RUNNING_SQL view . This is a very useful DB2 administrative view , presenting long – running queries. Under the hood , the LONG_RUNNING_SQL view joins some system snapshots. SELECT APPL_NAME,AUTHID,INBOUND_COMM_ADDRESS, STMT_TEXT, AGENT_ID, ELAPSED_TIME_MIN, APPL_STATUS, DBPARTITIONNUM FROM SYSIBMADM.LONG_RUNNING_SQL ORDER BY APPL_NAME Check the APPL_STATUS... Read more →

How to write a DB2 loop with INSERT

Question: I’d like write a sql statement to loop through an INSERT statement and increment with a count. The purpose is to create some test tables for load testing. Answer: It is possible to create a loop in DB2 which loops through an incremental INSERT. This is a basic example, which can be customised for your purposes.Note the use of ATOMIC. The purpose of ATOMIC is to rollback before the call is passed back to the requestor, if there is a problem. In this example the CNT variable increments at every INSERT, up until it is under 100000. db2 “CREATE... Read more →

The great Windows or Linux debate for DB2 LUW

I manage large DB2 servers on Windows and Linux , so I’m confident both OS platforms can support installations with large data sets and high end throughput. But believing DB2 will perform exactly the same on both Windows and Linux without some analysis is a recipe for problems. There are some considerations when planning switching or deciding between DB2 or Linux or Windows. Scriptability of DB2 on Linux versus Windows. If migrating from Linux – you may have a large shell script library which manage Linux and commit the flow control . There may be considerable work to adapt these... Read more →

DB2 read-only clause and performance

Question: Why do some sql queries use a read-only clause? Is there any performance gain by using the read only clause? Answer:Sometimes you will see a sql statement such as : Select name from mytable for read only The purpose of using the read only clause is to make the cursor unambiguous. When the read only clause is used , positional updates cannot occur By unambiguous I mean : Normally the SELECT is turned into a cursor. If the cursor is explicitly defined as a READ ONLY – “blocking” is used. “Blocking” means the DB2 server sends groups of rows... Read more →