If you will be in the Columbus area, we will be hosting a DB2 Performance Workshop with Optim Performance Manager (OPM) May 12th in the IBM Office. To register, please email jkoblen@us.ibm.com
Optim Performance Manager Extended Edition for DB2® on Linux®, UNIX®, and Windows® is the latest offering from IBM to help you resolve emergent database performance problems before they affect the business. In this workshop, you will learn about the new capabilities provided in the product,including a powerful Web-based user interface with detailed diagnostic displays for problem analysis, out of the box reporting for both post mortem analysis and proactive trend analysis, enhanced integrations to support end-to-end diagnosis
and tuning, and expanded capabilities for deep performance insights into Java™
and DB2 call level interface applications.
Here is a link to a recent developer work's article which gives an introduction to the OPM solution.
http://www.ibm.com/developerworks/data/library/techarticle/dm-1004optimperformance/index.html
Monday, April 19, 2010
Wednesday, December 30, 2009
New Year's resolutions ..well..add Database to that diet too.
Wondering what your new year's resolutions will be this year ? Is dieting on that list ? It's almost always on mine, or at least eating better. But, how 'about adding dieting for your database too.
Much like my appetite for too much food throughout the year, many of our z/OS -unix/linux and Windows applications also have a large appetite for volumes of data. Maybe mergers and acquisitions happened, maybe lack of (or a myriad) of data retention policies, or any number of factors had your data volumes growing at a rapid pace. Does all this data have you constantly purchasing CPU and capacity upgrades ? Backing up and recovery too much data ?
Is all that data really used, maybe it's time to ask your database to go on a diet.
When we eat less, those favorite jeans fit better.........and well, when our database is lean & mean ..those queries will run better too.
I've decided it may be necessary, so I put this video together to show how to put that DB2 z/OS data on a diet using IBM Optim. This video shows how to use the native z/OS version of IBM Optim, but you can do the same kinda stuff for distributed platforms with Optim too. Against any number of database platforms such as Oracle, SQL Server, DB2 LUW ... you get the picture.
IBM Optim will pull that fat off your database and make those queries and backups run mean and lean. Optim has the capability to
provide universal access to that data without even making it necessary to restore it. However, if you do decide to restore that archived data - you can place it back in the original source, or even a different database
platform...
Well, here is a quick tour of archiving (aka dieting) for DB2 z/OS. You can also view the video on Channeldb2.
Still need more reasons to think about dieting (I mean archiving) ? how 'bout
–Improve performance
–Control data growth, save storage
–Support retention compliance–
Enable application retirement–or
Streamline those waistlines ...I mean streamline those upgrades. A Peoplesoft upgrade can feel a whole lot better after an Optim Archive diet.
You can find more information on our Optim Library : http://www.informationmanagementrequest.com/mk/get/OPTIMLIBRARY
Much like my appetite for too much food throughout the year, many of our z/OS -unix/linux and Windows applications also have a large appetite for volumes of data. Maybe mergers and acquisitions happened, maybe lack of (or a myriad) of data retention policies, or any number of factors had your data volumes growing at a rapid pace. Does all this data have you constantly purchasing CPU and capacity upgrades ? Backing up and recovery too much data ?
Is all that data really used, maybe it's time to ask your database to go on a diet.
When we eat less, those favorite jeans fit better.........and well, when our database is lean & mean ..those queries will run better too.
I've decided it may be necessary, so I put this video together to show how to put that DB2 z/OS data on a diet using IBM Optim. This video shows how to use the native z/OS version of IBM Optim, but you can do the same kinda stuff for distributed platforms with Optim too. Against any number of database platforms such as Oracle, SQL Server, DB2 LUW ... you get the picture.
IBM Optim will pull that fat off your database and make those queries and backups run mean and lean. Optim has the capability to

Well, here is a quick tour of archiving (aka dieting) for DB2 z/OS. You can also view the video on Channeldb2.
Archiving DB2 data with IBM Optim z/OS from jankobl on Vimeo.
Still need more reasons to think about dieting (I mean archiving) ? how 'bout
–Improve performance
–Control data growth, save storage
–Support retention compliance–
Enable application retirement–or
Streamline those waistlines ...I mean streamline those upgrades. A Peoplesoft upgrade can feel a whole lot better after an Optim Archive diet.
You can find more information on our Optim Library : http://www.informationmanagementrequest.com/mk/get/OPTIMLIBRARY
Thursday, June 4, 2009
New JDK version to enhance performance of SAP on zLinux
SAP and IBM announced a new J9-based JVM with enhanced memory management for the SAP NetWeaver stack. SAP Note 1338196 has the details on the general availability (GA). For Linux on System z, the following are now supported:
Per the SAP Note 1319038, you can download the code here and the aforementioned note has the technical details for installation:
https://www14.software.ibm.com/webapp/iwm/web/preLogin.do?lang=en_US&source=swg-ibmjavaisv
The announcement details stated " The J9-based JVM comes with an enhanced memory management for the SAP NetWeaver Java stack complemented by new efficient Garbage Collector policies. The resulting benefits in SAP Java environments are more efficient server utilization and more consistent response times. Internal and customer tests in SAP NetWeaver environments have shown reduced CPU utilization while keeping full compatibility with the IBM Classic VM for Linux on System z."
Using it yet ? Post your comments and feedback.
- SAP NetWeaver 2004 starting with SP22
- SAP NetWeaver 7.0 (2004s) starting with SP15,
- SAP NetWeaver 7.0 (2004s) EhP 1 starting with SP3
- SAP NetWeaver 7.0 (2004s) EhP 2 and future EhP’s
Per the SAP Note 1319038, you can download the code here and the aforementioned note has the technical details for installation:
https://www14.software.ibm.com/webapp/iwm/web/preLogin.do?lang=en_US&source=swg-ibmjavaisv
The announcement details stated " The J9-based JVM comes with an enhanced memory management for the SAP NetWeaver Java stack complemented by new efficient Garbage Collector policies. The resulting benefits in SAP Java environments are more efficient server utilization and more consistent response times. Internal and customer tests in SAP NetWeaver environments have shown reduced CPU utilization while keeping full compatibility with the IBM Classic VM for Linux on System z."
Using it yet ? Post your comments and feedback.
Monday, June 1, 2009
IBM Optim Technology Workshop in Cleveland, Pittsburgh and Columbus
Just a short post for the upcoming IBM workshop in the Great Lakes area.
IBM Optim Technology Workshop in
Pittsburgh on June 9 http://events.linkedin.com/IBM-Optim-Proof-Technology/pub/64145
Cleveland June 10 http://events.linkedin.com/IBM-Optim-Proof-Technology/pub/64142/
Columbus June 11 http://events.linkedin.com/IBM-Optim-Proof-Technology/pub/64148
Hope to see you there......
Heartbroken in Cleveland after Cavs loss.
IBM Optim Technology Workshop in
Pittsburgh on June 9 http://events.linkedin.com/IBM-Optim-Proof-Technology/pub/64145
Cleveland June 10 http://events.linkedin.com/IBM-Optim-Proof-Technology/pub/64142/
Columbus June 11 http://events.linkedin.com/IBM-Optim-Proof-Technology/pub/64148
Hope to see you there......
Heartbroken in Cleveland after Cavs loss.
Friday, May 29, 2009
A couple of things about DB2 STMM and zLinux
It's been quite some time since I've posted to this blog. There are so many good resources for DB2 information today that I've been struggling to find a topic I felt was worthy of posting. Maybe, since I live in Cleveland the only thing 'worthy' of blogging about may be Lebron James. But, I'm writing once again on Linux or zLinux for two reasons. One, to share with you some information I've acquired while assisting customers with DB2 and second, to get your feedback on your experiences.
I thought I'd post a collaboration of the information I've used while researching a few issues that some customer's have been experiencing. All of these have had something to do with either (STMM) DB2's Self Tuning Memory Manager or Linux (with a focus on zLinux), or both.
First, lets start with some clarification on the limitation of DATABASE_MEMORY=AUTOMATIC on some Linux kernels.
1) Some older linux kernels had limitations in the capabilities for supplying adequate OS information for STMM. Thus, the database_memory setting reverts from automatic to computed on these platforms. If you are on SLES10 SP1 or RHEL5 or higher, then this limitation has been removed.
The details of this limitation are documented in the InfoCenter at:
http://tinyurl.com/nnzmdh
OK, so your on one of the linux distributions and kernels that support automatic, but, you are still experiencing issues.
2) On Linux platforms there were several APARs involving STMM that were fixed in FP2 & FP3, so I'd recommend getting to FP3 if you are planning on using STMM on Linux.
Here are a few to note, but you can find all related APARs by searching for 'Linux+STMM' from the DB2 APAR Search.
3) Another common question or problem is the encounter of 'Unable to find donor' . This technote describes how and why that may be happening:
http://tinyurl.com/nps4bp
4) On zLinux, DIO (Direct I/O) is not the default setting. This is because the current support for zLinux requires SCSI over FCP due to alignment. This requirement is documented at :
http://www-01.ibm.com/support/docview.wss?uid=swg1LI72705
In one of my older posts, I noted DIO as one of my top ten tuning tips.
http://justdb2chatter.blogspot.com/2008/03/linux-on-z.html
If you are going to use STMM, I would recommend that you use DIO. Actually, DIO is a best practice even if your not using STMM. To learn more about how DB2 and the linux file system cache relate, this article on developerworks does as nice summary.
http://www.ibm.com/developerworks/data/library/techarticle/dm-0509wright/
You can find additional tuning tips for zLinux at:
http://www.ibm.com/developerworks/linux/linux390/perf/tuning_pap_database.html
Lastly, if your using STMM on a zLinux platform I'd like to hear about your experiences, the good , the bad and the ugly. Especially if you are using or testing Collabrative Memory Management (CMM) for z/VM with DB2 9.5 STMM enabled. Comment or email me, I'm interested.
I thought I'd post a collaboration of the information I've used while researching a few issues that some customer's have been experiencing. All of these have had something to do with either (STMM) DB2's Self Tuning Memory Manager or Linux (with a focus on zLinux), or both.
First, lets start with some clarification on the limitation of DATABASE_MEMORY=AUTOMATIC on some Linux kernels.
1) Some older linux kernels had limitations in the capabilities for supplying adequate OS information for STMM. Thus, the database_memory setting reverts from automatic to computed on these platforms. If you are on SLES10 SP1 or RHEL5 or higher, then this limitation has been removed.
The details of this limitation are documented in the InfoCenter at:
http://tinyurl.com/nnzmdh
OK, so your on one of the linux distributions and kernels that support automatic, but, you are still experiencing issues.
2) On Linux platforms there were several APARs involving STMM that were fixed in FP2 & FP3, so I'd recommend getting to FP3 if you are planning on using STMM on Linux.
Here are a few to note, but you can find all related APARs by searching for 'Linux+STMM' from the DB2 APAR Search.
http://www-01.ibm.com/support/docview.wss?uid=swg1IZ18366
http://www-01.ibm.com/support/docview.wss?uid=swg1IZ26166
3) Another common question or problem is the encounter of 'Unable to find donor' . This technote describes how and why that may be happening:
http://tinyurl.com/nps4bp
4) On zLinux, DIO (Direct I/O) is not the default setting. This is because the current support for zLinux requires SCSI over FCP due to alignment. This requirement is documented at :
http://www-01.ibm.com/support/docview.wss?uid=swg1LI72705
In one of my older posts, I noted DIO as one of my top ten tuning tips.
http://justdb2chatter.blogspot.com/2008/03/linux-on-z.html
If you are going to use STMM, I would recommend that you use DIO. Actually, DIO is a best practice even if your not using STMM. To learn more about how DB2 and the linux file system cache relate, this article on developerworks does as nice summary.
http://www.ibm.com/developerworks/data/library/techarticle/dm-0509wright/
You can find additional tuning tips for zLinux at:
http://www.ibm.com/developerworks/linux/linux390/perf/tuning_pap_database.html
Lastly, if your using STMM on a zLinux platform I'd like to hear about your experiences, the good , the bad and the ugly. Especially if you are using or testing Collabrative Memory Management (CMM) for z/VM with DB2 9.5 STMM enabled. Comment or email me, I'm interested.
Wednesday, January 7, 2009
IBM Software Support Toolbar - great for quick DB2 searches
I recently spoke with a customer who was not happy with our DB2 documentation. I'm actually pretty fond of the Information Center myself, I find the search capabilities nice. However, if you need to search for a DB2 topic and you aren't sure where to start. The IBM Software Support Toolbar is a great gadget. I've added mine to my FireFox browser and it's located right underneath my Google toolbar. I use both of these toolbars quite often. The IBM toolbar will allow you to search through all documentation, APARS, troubleshooting docs and learning docs with one click of the button. It also provides quick links to valuable support resources such as, Passport advantage and ESR (Electronic Service Requests).
It's a handy tool in a day and age of information overload. I find the toolbar a valuable asset. The toolbar can be used for other IBM support products, such as Tivoli, Rational, Websphere and Lotus. I've focused my conversation about the Information Management since DB2 is where I tend to spend most if not all my time.
If your interested in obtaining the toolbar you can download it for free from:
http://www-01.ibm.com/software/support/toolbar/index.html?ibmsst=ibmTbMenu
It's a handy tool in a day and age of information overload. I find the toolbar a valuable asset. The toolbar can be used for other IBM support products, such as Tivoli, Rational, Websphere and Lotus. I've focused my conversation about the Information Management since DB2 is where I tend to spend most if not all my time.
If your interested in obtaining the toolbar you can download it for free from:
http://www-01.ibm.com/software/support/toolbar/index.html?ibmsst=ibmTbMenu
Saturday, November 1, 2008
DB2 8 goes out of support soon.
I've recently visited several customers who still have or are using V8 in their distributed environments. I thought I'd post this very short note as a reminder that DB2 V8 will be going out of support in April (09). You can view the details and announcement here:
http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?subtype=ca&infotype=an&appname=iSource&supplier=897&letternum=ENUS907-125
It seems that date is arriving much sooner and faster than some expect. But getting to V9 or V9.5 is easier than you think. There is alot of good information available to help you, and the migration portal is a great place to start.
http://www-01.ibm.com/support/docview.wss?rs=73&uid=swg21200005
This Chat with the Lab webcast on migration can also assist you:
http://www.eseminarslive.com/c/a/Database/IBM091908/
So, take a look at the material and prepare to upgrade if you haven't already started.
http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?subtype=ca&infotype=an&appname=iSource&supplier=897&letternum=ENUS907-125
It seems that date is arriving much sooner and faster than some expect. But getting to V9 or V9.5 is easier than you think. There is alot of good information available to help you, and the migration portal is a great place to start.
http://www-01.ibm.com/support/docview.wss?rs=73&uid=swg21200005
This Chat with the Lab webcast on migration can also assist you:
http://www.eseminarslive.com/c/a/Database/IBM091908/
So, take a look at the material and prepare to upgrade if you haven't already started.
Subscribe to:
Posts (Atom)