Among the 'recommended practices' that I often present on DB2 for z/OS stored procedures is this one:
- No more than 512 SP's in a WLM
Let me explain why I recommend this. It's actually at the bottom of the list, and that's because it doesn't come up that often. But it has, and when it does, it can cost in I/O. DB2 has a Language Environment table of load modules in each stored procedures address space. For stored procedures defined STAY RESIDENT YES, we only have room for 512 load modules in that table. A load module has to be in the table in order for DB2 to invoke it. So, starting with the 512th, we'll delete it from the table after we call it, even if it's STAY RESIDENT YES. And come to think of it, we have separate tables for TYPE MAIN and TYPE SUB.
So to be completely accurate, the recommendation could actually say something like this:
- No more than 512 different load modules for STAY RESIDENT YES SP's in a WLM application environment, that are all either PROGRAM TYPE MAIN or PROGRAM TYPE SUB and invoked during the lifetime of a single instance of a WLM-SPAS.
For that last bit, remember that different invokers of a stored procedure that end up classified in different WLM enclaves will not have their SPs run in the same instance of a WLM-SPAS.
What's a WLM-SPAS? It's what I use to abbreviate a "WLM-established stored procedures address space".
And this post has motivated me to get a more recent copy of my stored procedures recommended practices presentation out online![Read More]
Also among the 'recommended practices' that I often present on DB2 for z/OS stored procedures is this one:
- Don't call the metadata stored procedures
Many invocations of DB2 for z/OS stored procedures come from a Java(TM) or a CLI application. The software stack for these programs accessing DB2 for z/OS is through a "driver" program. These driver programs have SQL packages bound to DB2 for z/OS, and in the case of the application invoking a stored procedure, there is a fair amount of code executed in the driver program.
For a CLI program (the term CLI is often used interchangeably with ODBC) -- this is usually something running from a Microsoft(TM) application accessing DB2 for z/OS. The DB2 connect software that includes the driver for DB2 for z/OS has some smarts in it so that if the application is coded using incorrect data types for the stored procedure being invoked, the driver recovers and invokes the SQLPROCEDURECOLS metadata stored procedure on DB2 for z/OS to find out what the data types are and then re-sends the stored procedure call to DB2 for z/OS. Yes, you got it right, this means that a poorly coded application can invoke 3 stored procedure calls for every SQL CALL it's trying to do -- one to the original SP, one to SYSIBM.SQLPROCEDURECOLS, and then again to the original SP with the correct parm types! How do you recognize this? Well, you could run a client-side DRDA trace and it will show up there. Or you can look at statistics at the server. Or you can set the value DESCRIBEPARAM=0 in the db2cli.ini file on the client, and let the applications get the error SQLCODE -301 because now the driver won't do the metadata PS call and instead will let the application fail due to using the wrong datatype. Same result if you issue a -STOP PROCEDURE (SYSIBM.SQLPROCEDURECOLS) ACTION(REJECT) command on the DB2 for z/OS server.
For a Java(TM) program, the current driver is the DB2 Universal Java Driver, and it will not invoke the metadata stored procedure. So this is an excellent reason to switch to the current driver, because the older version of the driver went through the CLI code path and had the same problem as described above.
Note that if you invoke a stored procedure from the command line (the CLP), that code will always invoke the SQLPROCDURECOLS stored procedure since the command line doesn't provide anything for what data type the arguments are.
Now, if you are stuck with a CLI program that you can't modify, what can you do to improve the performance of SQLPROCEDURECOLS? Well, APAR PK57017 just shipped which reduces the size of the package for this stored procedure, so you can free up some EDM pool usage and get a small CPU usage improvement. You can also be sure you run RUNSTATS so that the data access for this SP is the most efficient it can be. I have also heard rumors of some customers creating additional indexes on the tables used by SQLPROCEDURECOLS, but I don't have any specifics on that, sorry.[Read More]
Too many blog posts start with "It's been awhile since I've posted here", so I'm gonna skip that part and pretend that we've been talking regularly all along, OK ?
Firstly, yes, I'll be at IOD
this year! I've missed a couple due to being completely heads-down on helping get this cool new product out, so it will be great on a personal and professional networking level to reconnect with many of you. I'm there to help represent an IBM product called IBM Content Analytics
affectionately known as "ICA". It's part of our Enterprise Content Management suite. ICA integrates very well in an enterprise but at its core it is a standalone product - no prereqs. I hope that we get a chance to chat about it if we run into each other at IOD, I promise to keep it fun and friendly. :-)
The value proposition for IOD is one of unlocking insight with business value from your text content. Of course there is amazing technical capability behind it, and that's where my time is spent. Remember those old ads for the Evelyn Woods Speed Reading
courses (kinda funny that it's still around, huh!)? Conceptually, it's like this ICA system knows how to do that, because it can just go and "read" anything and everything you've got, and then tell you what the important things are across the whole smash of documents. You aren't searching, you're being presented the hot terms, words, phrases, words that come from a list you provide, and
how they are trending over time (you get to choose the date for analyzing trends - received date? incident date? date extracted from text? hire date? birth date? -- get the picture?). ICA also shows you how other terms, phrases, words correlate to what you've already identified. Right away out of the box, ICA will show you insight -- and it connects to all different enterprise sources
. We've got a great suite of built-in capability for text analytics and intuitive visualizations for your insights. And then on top of that, ICA is completely customizable, too - make it find the insights/concepts you are interested in, display them how you need to, get instant results for a document, export analytics results to your favorite reporting tools, and oh, so much more!
My technical lead focus on 2010 has been on the ICA tooling and customization capabilities - how to add new custom text analytics, how to use the LanguageWare Resource Worbench
to customize ICA, and how to make sure the sources of text are represented the way you want them. In that vein, I led a team of brilliant capable students in an Extreme Blue
project this summer who really helped us to show how you can unlock the value in your data using this platform. And I've been having a lot of geeky fun designing and giving demos to lots of different clients.
Come learn more if you'll be at IOD. It's easy to search this ECM roadmap
for Content Analytics sessions, and that's where you'll find me hanging around, plus I'll be at the Expo pedestals. Or let's catch up over a beverage - it's been too long!
I heard an interesting story on the news last week, about how the individual states of the U.S. were graded on how they use information. The state I live in, California, got a C+. How can this be, with our advanced technologycenters in Silicon Vallley?
I found the article online here and found some interesting things, although nothing specific about California.
The article says:
When all is said and done, a state’s skillwith information is found at the intersectionof three distinct operations: the willingnessto share data, the capacity to generategood information, and the ability toget those who should use the data to do so.
Well, that sounds a lot like stuff that I have talked about when describing IBM's Information on Demand strategy. Is your organization good at doing this? I particularly noted the last point in the article, because some of the states complain that their legislators just aren't interested in using the data! Maybe we information professionals have to make that easy (and fun?) to do.
What about the highest-graded states? The article had this to say about one of them:
In Washington State, Governor ChristineGregoire held a series of town hallmeetings on the budget to communicate resultsto citizens and follow up on the budgetarypriorities she had previously establishedwith much citizen input. “We wantto give concrete information about whethera difference has been made or hasn’t"
Yep... this is what everyone wants to know. What did we say we'd do? Did it make a difference? In fact, I've been trying to get this type of information from my financial analyst for some time!
What about states that were graded worse than California?
Some state employees in Rhode Island arestill operating with typewriters—electric, ofcourse, but still a far cry from the ability toshare information in a database. NewHampshire has such weak data-sharing systemsthat it doesn’t know how much itspends each month—kind of like an averageJoe who’s lost his checkbook.
At the opposite end of the spectrum, there’s Wyoming. Itstransportation department has linked geographicinformation systems to financialsystems and now knows with exact specificityhow money is being spent, down to thecost of the salt used between each milemarker on the state’s snowy roads.
OK, well, perhaps that is an example of too much information! :-)[Read More]
We are often asked where to find the sample files for our text analytics tutorial. It's a small set of IBM quarterly reports. I've uploaded them here, just click on that to download them. Happy coding!!
When I describe native SQL procedures in DB2 9 for z/OS, I often hear variations of these types of questions:
- Doesn't the external WLM-managed infrastructure provide some throttling of stored procedures? What's going to happen when this is gone?
- Can DBM1 handle the same amount of concurrent stored procedures as multiple WLM-SPAS?
- User routines only use below the bar storage, so how much below the bar storage is available in DBM1 for these native SQL procedures?
In order to answer this, I have to explain a little bit about how DB2 handles native SQL procedures. They are simply packages, with "runtime structures" for the SQL statements to be executed. So, when you invoke a native SQL procedure, DB2 finds and loads the package and executes the statements.
In contrast, an external stored procedure with SQL needs a complete language environment for the user program, and then that external program comes back to DBM1 to get its package loaded and SQL statements executed. That's what needs to be "throttled" - the external program execution environments and their associated TCBs. When an incoming stored procedure request is queued for WLM, the DB2 thread is suspended in DBM1. Many customers have experienced delays and DBM1 storage problems when their WLM goals weren't adjusted properly and the queued requests built up. The solution is to either adjust the WLM goals, or else adjust the limit on DB2 threads (local and/or distributed).
With native SQL procedures, the thread will just switch packages when the call statement is processed and run the procedure - no queuing. The storage used for the local variables is above the bar and managed with efficient algorithms. The maximum concurrent first-level native SQL procedures is effectively the same as your setting for maximum DB2 threads. (What I mean by first-level is that a native SQL procedure may have a nested call to another native SQL procedure, so the actual number of concurrent native SQL procedures may be even higher).
So, I guess the way I'd answer the questions is:
- Yep. When it's gone, SPs will run much more efficiently
- Yep - in fact likely more
- n/a - SQL procedures aren't "really" user routines - they are a pre-defined set of SQL statements, and they don't use below the bar storage
Of course I recommend that you test your native SQL procedures in your environment and measure for yourself, and do capacity planning based on the results of your testing. Native SQL procedures will use some DBM1 storage, after all, and how much depends on what statements and what variables are used in the program.
Oh, and if you didn't recognize it, the "What, me worry?" is a reference to the signature quote from Alfred E. Neuman. It's more than a little tongue-in-cheek.[Read More]
As many of you know, I changed jobs early in 2008 to switch my developer focus from DB2 for z/OS over to unstructured text technologies. Since many DB2 folks are heads-down in structured data, that whole "content"side is a bit of a mystery. Sure, you know about search, and you likely have a vague understanding of what it means to have a text index supporting keyword search. But really, there's so much more...
I've been learning a ton about all of this, which is to be expected after over a year (!) in this job. I have some terrific colleagues in text analytics, in research, development and services, who continually amaze me with their breadth and depth of knowledge, as well as their passion for the topic and their eagerness to help customers.
I happen to love linguistics, so this job is a great fit for me. I love to read, and the turn of a phrase in a book or a song lyric brings me joy. I like to think about the best way to phrase things and ways to interpret sentences. The more I interact with non-native English speakers, the more I appreciate both the beauty and the limitations of language, and the inherent difficulties in both generating and understanding sentences. I truly enjoyed all my study of French in school, too. It's always been such an intellectual snobbery to say something like "the French have a word for that", but anyone who knows more than one language knows it to be true -- language translation is never exact and concepts cannot always be expressed well, even in one's native language. Bottom line is that it's so interesting for me to dig into and help shape the technology and rules around extracting meaning from unstructured text.
Last month I was talking with a long-time friend and colleague who was here with her company at the IBM Silicon Valley Lab for a technology briefing. She and I have had several conversations at conferences over the years on topics like O-O databases, Java, XML, as they were emerging towards mainstream over the years. In the briefing, we talked a bit about unstructured data in the context of the Information Agenda, and one of their company's thought leaders said that unstructured data inclusion is implied. Cool, but, um, how exactly? Their (very reasonable) response when I probed a little further was that they needed to hook up with the business guys on that. YES! That's where I think we all absolutely have to start -- what is the unstructured data, and what questions do you need answered from it for business value. Then we go into more of the logistics around that.
Specifically, just this week I've worked on a couple of items that can help me bring some meaning to what text analytics is all about to folks who haven't been exposed to it deeply. The first was working on a report for a well-known analyst group, where we describe our information access technologies and offerings for unstructured data. And the second is a new offering that I hope will become available soon, to help quantify needs and specific business value that can be derived from unstructured data. If you are curious about any specifics, a great place to start digging into and even playing with some text analytics technology is the LanguageWare capabilities, system text and UIMA.
If you're interested in this kind of stuff, please let me know, or contact your local IBM rep and ask them (and tell them to ask me if they want a starting contact!) I'm passionate and eager to help! :-)[Read More]
Even though I don't work daily on DB2 any more, I thought this might be a helpful post.
Last night I was online, shopping coincindentally on Cyber Monday, for presents for both my Dad's birthday and my nephew's birthday.
A work colleague happened to catch me online to ask some questions. Now, these answers are not from a book, just what I said right out of my head -- JUST as I would if you and I had this conversation. Here is our Lotus Sametime conversation, verbatim, no editing. updated: one minor edit marked by  -- OK, and I also put in the html tags and made the requester anonymous, of course. So -- sure, I could have said more about the virtues of data sharing and WLM. They are many! Forgive me, it was nighttime and I wasn't in a "marketing" mood.
In any case, please enjoy the truthiness of this exchange.
Dec 1, 2008
8:25:22 PM coworker Hi, do you have time to give me a 101 lesson on db2 z stuff?
8:28:11 PM me right now?
8:28:19 PM coworker yes
8:28:38 PM me what do you need to know?
8:29:07 PM coworker ok here come the stupid questions...
8:29:10 PM what is data sharing?
8:29:58 PM me it's 2 DB2 subsystems that share one set of data, for availability and scaling beyond one machine
8:30:06 PM 2 or more that is
8:31:02 PM coworker ok so that leads me to another question, can you explain subsystems, lpars ? what is it like in windows world?
8:31:46 PM me a subsystem is a DB2 installation
8:32:18 PM an lpar is a "logical partition", kind of like a VM on windows...
8:32:59 PM coworker can you install anything else on a subsytstem?
8:33:12 PM how do subsystems and lpars relate to each other?
8:34:34 PM me a subsystem is just db2's code, it runs in an operating system which is in an LPAR. other programs (like IMS and websphere) can be installed in that same operating system instance in that same LPAR
8:36:49 PM coworker oh oi see, i'm getting confused with the subsystem term to mean a z/OS OS term
8:36:50 PM got it
8:36:57 PM so tell me about hypersockets?
8:37:56 PM me they are a "fast pipe" for TCP/IP between LPARs on a physical z box - like a 1GB network but faster (don't know the exact numbers)
8:39:10 PM coworker so is this DRDA or no?
8:39:15 PM or using db2 connect
8:40:30 PM me DRDA communication messages travel over over TCP/IP, so if both sides are in LPARs, then they can go over the hypersockets - such as Linux on z on one side in an LPAR, and DB2 z on the other side in an LPAR
8:41:27 PM coworker so when 2 lpars communicate they do not use TCp/IP, they use hypersockets
8:41:30 PM is this correct?
8:42:32 PM me they do use TCP/IP, which travels over the hypersockets as a physical path - like a network wire
8:42:53 PM now... 2 datasharing members don't use either, if that's what you're getting at
8:43:12 PM they use the coupling facility to communicate
8:43:21 PM coworker is this built into z/OS or db2?
8:43:26 PM me z/OS
8:43:26 PM coworker the copuling faciloity
8:43:29 PM i see
8:43:50 PM me it's a separate feature of z - parallel sysplex, which is what datasharing is built on
8:44:25 PM coworker how come a customer would want to use data sharing? benefits? cons?
8:45:26 PM me benefits are availability, since all of db2 maintenance can go on one member at a time so they can keep their data available
8:45:49 PM also scalable, since the number of threads to connect to one db2 is limited, the other threads can go to another db2 -- they share all the data
8:45:51 PM coworker o i c
8:46:02 PM ok got it
8:46:08 PM me good!
8:46:13 PM coworker wait one more question hahaha
8:46:15 PM me k
8:46:17 PM coworker tell me about workload mgmt
8:46:22 PM tha'ts a z/OS thing right?
8:47:05 PM me yes - it helps balance when there's more than one subsystem on the LPAR, also helps balance priority between different db2 threads and even helps choose which data sharing subsystem gets a connection based on how busy the systems are
8:47:57 PM coworker how come this is such a big advantage over windows let's say
8:48:06 PM doesn't all operation systems do this
8:48:56 PM me no - windows doesn't let you set the priority of different programs/users/etc (as far as i know)
8:50:52 PM coworker yes you can via task manager
8:50:55 PM although not sure how smart it is
8:51:15 PM what makes z/os workload mgmt so sophisticated
8:51:21 PM i'm just wondering what's the big deal
8:52:28 PM me it's got a pretty complex set of rules and levels of priority, frequent sampling, decaying priority over time, different rules for different times of day, etc. That said, I didn't know it was considered that big of a deal
8:54:30 PM coworkerhahahaha
8:54:36 PM thanks for your db2 z/os 101
8:54:43 PM me any time[Read More]
I know that many of you are getting ready to attend the IOD conference
next week in Las Vegas. Alas, I will not be there, but that doesn't stop me from giving you advice about how to get the most out of a conference. After all, you are spending your time there, and your company's money, so you might as well make the most of it.
Earlier this month, I was fortunate to attend the Grace Hopper Celebration of Women in Computing in Keystone, Colorado. I was notified of my participation only two weeks before, and I wasn't aware of this conference at all. I made my reservations and printed out the conference agenda, which I reviewed on the plane on the way there.Most conferences give you access to the agenda ahead of time, and it's a good idea to print that out and have a look at it.
As for me, after one quick pass through it, I started by putting a "dot" next to anything I might be interested in, and a "star" next to anything that was a 'must see'. Then I started to notice some trends emerging. I came up with these goals for the conference, based on what was in the agenda:
- meet other IBM women (this was, after all, a Women in Computing conference, and IBM was sending me)
- text analytics technologies
- women in technology issues, including attracting more students
- social networking and collaboration
- cool stuff other than the above
To that end, I circled the name of any IBMer, and labeled each of the rest of what I noted as one of the above. Now, I realize, this approach might seem a bit, well, organized -- particularly for me. But it really worked, particularly to keep my attention and also keep a balance of different topics, as well as provide a tiebreaker when there were multiple sessions at the same time - I just look at the balance of the other sessions I've been attending.
Another thing you want to do while you're there is talk to other people about what they've seen, what they are going to see, etc. It's so easy to miss something or misinterpret something in an agenda.
If you keep a good balance between "use right now" and "good to know for the future" topics, you can help keep your brain from overflowing - that's always a danger at a multi-day conference![Read More]
I've received a few questions about text search in DB2 9 for z/OS lately, so I thought I'd share the basic information here.
Prior to DB2 9, there was an offering called "DB2 Text Extender". This was an early attempt at text indexing that runs on z/OS. It is dependent on some z/OS code called "Text Search". Later on, the team that worked on the extenders also released something called "DB2 Net Search Extender", aka NSE. DB2 for Linux, Unix, and Windows had a significant upgrade with NSE, but that same upgrade was not shipped for DB2 for z/OS. So DB2 for z/OS customers have not had a significant upgrade to text search since "DB2 Text Extender" in DB2 V7.
In DB2 9, there is a completely new text search solution. Text search is provided by a built-in engine function called CONTAINS. This solution requires an external text search server that runs on a Windows or Linux operating system, which is provided as part of the DB2 Accessories suite. It is not a direct replacement of Text Extender function, so applications and administration policies need to be updated for this change. The best source for detailed information about this is in the information center topic Administering IBM text search for DB2 for z/OS
. And I have an earlier blog entry
about the announcement.
What about the DB2 Text Extender? It's not available in DB2 9 for z/OS. That means you won't see an equivalent for DB2 V8 FMID JDB881C. And you won't need IBM Text Search FMID-HIMN230 in DB2 9 for z/OS either, because that was a prerequisite for the DB2 Text Extender.
The above is all specific to DB2 for z/OS. DB2 for LUW 9.5 FP 1 ships the same Omnifind text search server, and DB2 LUW still ships support for their NSE as well.
Now that we've got that all cleared up :-), next we'll evaluate some use cases for text search applicability.[Read More