Below are some of our most asked questions concerning the TurboTune mainframe optimization process and capabilities .
Mainframe Optimization Questions
Does TurboTune have Cooties?
This is by far the #1 question asked by our customers. Critical Path Software does NOT have cooties. We only use IBM vanilla recommended improvements stemming directly from the IBM Redbooks. Don’t let anyone tell you our recommendations are without merit or background. It’s simply not true.
Is the TurboTune mainframe optimization process safe?
Absolutely! Critical Path Software has NEVER caused a contention issue in any environment. Our recommendations are tested in development driven by the comfort level and standard procedures set by each individual client and partner. If it doesn’t work, you don’t utilize it. It’s that simple.
What is TurboTune?
The TurboTune process compares files against our database of “best practices” for file definition and manual analysis developed over nearly a 40 year time line in the mainframe space. Our goal is to ensure maximum mainframe optimization and minimized resource consumption resulting in lower operational costs. Our tag line “PERFORMANCE is PROFIT” is our mantra.
CPSI is the ONLY company that can accurately predict the percentage of improvement in critical area of the environment without adding software to your system or affecting logic. We are so confident in our process we get paid only after you begin to see cost reductions. This enables our clients to:
- Budget precisely the cost of our service
- Spend only when cost reductions are offsetting the expense
- More effectively negotiate hardware/software contracts
How current is TurboTune?
CPSI takes into account the evolution of the operating system and how these changes affect the underlying code. As IBM specifications change, TurboTune is updated to be in sync.
- Example: Old parameters are now obsolete. TurboTune recognizes these obsolete parameters. Z/OS 1.3, for example, makes changes to certain index levels. TurboTune has implemented this and has optimized examples from several customers.
CPSI’s proprietary software, TURBO, maintains pace with advances in IT technology:
SMS EXTENDED ADDRESSING:
TURBO makes recommendations to certain parameters, which must be changed. It makes adjustments, which must be made when you want to use extended addressing. Turbo uses algorithms to calculate new settings—it adds bytes, where necessary, to the key size. This is a precise process. It must be accurately determined when to use SMS Extended Addressing. TURBO Indicates when a file should be used with Extended Addressing. When Extended Addressing is utilized specific changes must also occur. TURBO identifies these changes.
SMS-VSAM UNDER THE COUPLING FACILITY:
SMS allows On-line and Batch to run simultaneously through cache sharing and other things. TURBO sets up cache buffers, calculates proper CI sizes and then adjusts, as necessary.
RLS – RECORD LEVEL SHARING:
TURBO identifies parameters that cannot be used with RLS. CI sizes are then adjusted accurately to ensure best performance.
TURBO allows overlay of track readings for large files. TURBO identifies parameter compatibility and which files can effectively be used with STRIPING.
DB2 V8 & V9 and DB2 BOOTSTRAP FILES:
TURBO has always handled DB2 System Files in conjunction with CICS. TURBO improves files that handle the processing of DB2, V8 and V9. TURBO reduces the I/O of DB2 files, Bootstrap files, DCA work files and CICS that handles DB2 On-Line.
DATA FILE COMPRESSION:
TURBO identifies files capable of utilizing compression. Errors in file compatibility may result in heavy resource consumption. Compression errors are primary culprits for high CPU costs. Cost benefit of CPU consumed to DASD saved using compression in error is often a key driver for savings.
TURBO calculates Index CI sizes for Z Computers. Z machines utilize a larger addressing scheme, more keys to run servers, C++, etc.
CPSI’s outsourcing service — This service allows a client to outsource VSAM maintenance to us. CPSI will complete an analysis, develop 100% accurate improvements, and rank file definition with all improvements incorporated. The client will test and install.
What problems does the TurboTune suite address?
CPSI Addresses two primary concerns for mainframe managers and their executive partners
- Excessive costs in the mainframe environment arising from resource consumption
- Poorly performing applications causing latency and stability concerns
- TurboTune may result in:
- Reduced need to add CPU engines or additional MIPS
- Reduced additional software and/or software licenses due to additional MIPS
- Reduced addition of DASD. DASD may be inexpensive however petabytes can add up
- Reduced personnel overtime and/or additional headcount due to slow processing
- Reduced outsourcing costs. Outsources want you to spend. Reduce your consumption and reduce your costs. PERFORMANCE is PROFIT!
- Reduced costly reruns due to poor response times
- Reduced downtime from on-line contention
- Faster online response time. This problem is amplified during peak hours
- Batch windows decreasing – online coming up early enough and on-time
- Reduced Service Level Agreements (SLA’s) in jeopardy.
The TurboTune analysis addresses everything above. Companies often do not realize how inefficient the I/O subsystem files have become. The I/O subsystem is not regularly serviced. There are millions of lines of code in any company consuming resources. This code will often run regardless of how poorly it is set up. Performance is critical to reduce consumption costs. PERFORMANCE is PROFIT!
When should LISTCAT be collected?
Run the LISTCAT off prime shift, if at all possible, and after CICS is taken down. Typically file size will beyond the email limits. CPSI provides the ability to upload directly and securely to our server.
- Note that some files will show up on one LISTCAT and not another. Temporary work files come and go. This is normal and does not affect the analysis.
What tuning recommendations do you provide?
TurboTune can determine and recommend several types of cost saving file changes:
- Calculate optimal FREE SPACE settings for each file
- Define and allocate space optimally
- Adjust buffer and CI sizes for CICS LSR buffers
- Remove obsolete parameters
- Change incorrect SHAREOPTIONS parameters
- Identify CATALOG errors
- Identify broken catalog entries
- Reduce excessive index levels
- Reduce excessive CI and CA splits
- Reduce excessive EXTENTS
- Reduce excessive EXCP’S
- Reorg recommendations
- Identify files approaching the 4GB limit
- Identify old unused files
- Identify correct ci sizes for current DASD type
How does TurboTune identify which jobs need to be tuned?
It is the files used by jobs that need to be tuned and not the jobs themselves. Tuning the files improves job performance. Using the TURBO RANKING software CPSI can identify the worst performing files. This allows CPSI to prioritize and recommend changes that will have maximum impact on the system. This ensures cost reductions begin immediately.
Does TurboTune estimate the potential improvements in response time or cycle time?
TurboTune estimates the percent reduction in batch wall clock time and the percent reduction in online response time.
Can TurboTune determine if free space is being used effectively?
TurboTune determines if the free space setting for a file is optimal. TurboTune evaluates free space usage requirements by analyzing the number of inserts, deletes, updates and retrieves to a file and the current split ratio(s) within each file. Based upon the current number of splits and EXCPs, an optimal size for Freespace will be determined.
TurboTune can determine if too much free space is defined. Once determined, TurboTune generates the changes to be applied to optimize the free space setting. TurboTune also identifies free space that is allocated and never used.
Can TurboTune identify the benefit of data striping?
Data striping is most effective in a heavy sequential read environment, although it can be used in an online environment. TurboTune examines the ratio of data reads to index reads. A high number indicates a good candidate for data striping. TurboTune determines if the current stripe count is over or under allocated and if striping is defined.
How does TurboTune address compression?
TurboTune identifies files capable of utilizing compression. Errors in file compatibility may result in heavy resource consumption. Compression errors are primary culprits for high CPU costs. Cost benefit of CPU consumed to DASD saved using compression in error is often a key driver for savings.
CPSI often finds that properly tuned VSAM files provide excellent performance and require less resource consumption. MIPS are more expensive than DASD. In most cases CPSI find tuning more cost effective and may recommend the client discontinue the use of software compression.
A problem we have with some applications is that they run re-orgs too frequently and waste CPU and time. Can you recommend how often a VSAM dataset needs to be re-orged?
Reorganizing of a VSAM file may simply cover-up the problem and allow for splitting to begin again.
TurboTune will reduce the amount of splitting by using better parameters. Files that experience heavy insert and delete activity are the best candidates for improvement. They are identified by our Turbo Ranking software. If your file has any clustering of insert activity, reorganization may do more harm than good. Future clustered inserts may cause splits again in the same places. You may be able to avoid some of these additional splits by performing frequent reorganization. We identify files that require more frequent re-orgs than others. TurboTune will resolve splitting and dramatically reduce the need for re-orgs. With speed, caching and read ahead algorithms of modern DASD subsystems, there will be little if any performance benefit from doing a re-org.
Can TurboTune effectively optimize our NSR/LSR buffers? (CICS & BATCH)
TurboTune will analyze NSR/LSR buffer pools and make recommendations to improve efficiency, if needed. TurboTune identifies files that should be isolated in private NSR buffering.