Every year IBM issues a press release touting its IT support for the U.S. Open tennis tournament in Flushing Meadows, N.Y. In past years, they have always talked about the Power processor technology and the benefits of System p and i, as well as Linux on i and p.
This year, their message is about going “green.” Surprise.
Not only that, but read into the details and you’ll discover that they’re only running Linux on Power now. No AIX. No IBM i (they weren’t running IBM i before, but still).
Timothy Prickett Morgan has a good, albeit short, piece, grousing about this U.S. Open development.
Considering that IBM doesn’t own a Linux operating system and has just rejiggered pricing on i 6.1 and AIX 6.1 to make them more competitive with Linux, the Grand Slam systems that IBM makes available to the U.S. Open, the French Open, and the Australian Open would seem to be a perfect place to show how a mix of AIX and i on Power6 iron is the best way to support modern Web applications. I mean, IBM does want to sell its own operating systems, right? And if AIX and i can’t–or won’t–compete against Linux on a level court with a study and straight net, well, what’s the point?
In past years, we’ve written about IBM running the U.S. Open on the Search400.com site. The backbone of the operation used to be System p5 550 servers running AIX and Linux partitions, with the AIX partitions running player search and feedback applications. Meanwhile, Linux partitions on the System p servers and two System i servers ran other Web-serving applications as well as applications at the event hosting the Web site’s scoring system.
Now, apparently, it’s all Linux. At least it’s still on Power, but like Morgan says, this would be a prime opportunity to show off AIX and IBM i. Instead, Linux takes over the No. 1 seed.
We recently received a reader question in response to a reader-submitted tip on using SQL on System i to color source code and inline comments. On August 27, 2008, Don sent us this inquiry:
This is in response to a tip published 05/2008 concerning applying color coding using SQL to update SRCDTA in your program source files. I tried it and got the message (STRSQL): Argument *N of function CONCAT not valid. Any advice on this?
I passed it on to the tip author, Narendra Devireddy who responded with this information:
Although I cannot confirm with the details given, it might be because Don’s source file length may be less than 112 (i.e. SRCDTA less than 100 bytes), so when he tries to use SUBSTR for 100 bytes it might be giving this error.
Please let the user know that in the second query ” UPDATE QTEMP/ALIAS# SET SRCDTA = SUBSTR(SRCDTA,1,4)||X’22’||SUBSTR(SRCDTA,6,94) WHERE
SUBSTR(SRCDTA,7,1) = ‘*'”
the second parameter of highlighted SUBSTR should be 6 less than SRCDTA length(i.e src file length-18, as 12 bytes are reserved for SRCDAT & SRCSEQ), and try it out
Using Source files of length 112 is standard these days for RPGLE programs, and therefore I have built the query to fit that standard. Please do let me know the result after the change.
We passed this back to Don and he let us know that in fact, the problem was resolved but through slightly different means:
Thanks for the timely response, the problem was actually that the SQL session attributes was set to “SQL Rules” = *STD instead of *DB2, after the change it worked fine. Thanks for the help.
As always, if you find something in a tip that doesn’t work for you please let us know. Also, if you have general questions there are many helpful people on IT Knowledge Exchange ready to answer your AS/400 questions. Additionally, Narendra recently submitted a follow-up tip for creation of multicolored comment lines that may be of interest.
Jon Paris and Susan Gantner have an interesting article this month in the IBM Systems Magazine about how some independent software vendors have been a little behind the eight ball in getting their applications ready for IBM i 6.1 (V6R1), the newest System i operating system, or at least, making it well known that they are.
Since V6R1 came out in the spring, the authors were a little surprised that the ISVs were dragging their feet on this:
Another thing that came up during our investigations was the number of companies that still refer to their 6.1-ready software as a beta. Considering how long ago i 6.1 was announced, and that it’s been generally available for several months, this was a real shock to us.
The article then gives eight tips for either figuring out whether the ISV’s apps are 6.1-ready, or encouraging them to do so. The steps include things like running ANZOBJCVN, checking out upgrade costs with the vendor, and contacting IBM ISV group.
Mid-deranged.com’s “article” is called “Egyptians More Advanced than Earlier Believed.” Turns out they were running a data center with an AS/400. But I’ve got some problems with this piece. Clearly it wouldn’t be an AS/400 they would be running, but rather a System/38 or System/36, right?
If your company is considering a switch from the iSeries to pSeries and you are skeptical, you may find a story that John M. Willis recently shared on his IT Management and Cloud Blog interesting (and even helpful to use in any discussions with CEO’s who want to “change” things). In, The Laughing Boardroom – IBM eServer iSeries, Willis shares the story of a utility company that had been running an AS/400 for 20 years and then converted to the pSeries.
The initial design took them from one production AS/400 (they called it a mainframe) to about 10 P5 physical boxes with about 50 lpars of which about 20 lpars eventually became the replacement solution. On the software side they went from a single box that basically handled everything to a very complex infrastructure.
He provides some very specific (and alarming) statistics about the numbers of additional personnel and equipment this change required. The post also includes an IBM video that appears to be “unavailable” (but it still plays the audio fine if you don’t mind the dark screen).
This is one example, but do you know of more? What kind of pressure have you received to switch to a different system? What statistics do you use to combat these requests? What’s the number one reason you hear from management to consider a move to a different system? Share your stories in the comments here or, if you have a long story, send me an email and we’ll consider publishing it as a stand-alone post.
- IBM designates PTF security group for i OS: For years, IBM has stuck System i operating system program temporary fixes (PTFs) into other groups. Now it has its own, which could help i users find the PTFs they need instead of searching endlessly among them all.
- Big Blue takes on Microsoft with Lotus: The headline in itself isn’t news, but the fact that IBM is pushing Lotus hard by including how many downloads users made and how it fared compared to Microsoft, is a good sign, according to David Vasta.
- Two System i management tips: Designating authorities for journaled objects and Using the system request menu.
- IBM announces replacement for i5/OS (not for real) : Mid-deranged.com’s fake, satirical story on IBM changing the System i operating system to VU/OS, with the VU standing for Virtual Universe.
Last week I spoke to Ross Mauri, the IBM general manager of Power Systems, who talked some about the IBM financial numbers that recently came out. Mauri did some bragging about how Power Systems, which IBM called the “converged System p” line in its release, was up 29% year-over-year. Although this number is compared to just the System p servers alone from the previous year, it was still a good quarter.
Mauri added that among Power systems, the high-end was up 21% and the midrange was up 68%. Well, what does that mean?
Typically IBM thinks of the 590 as the high-end, the 570 as the midrange, and the 550 and below as small. But the way in which IDC, the Framingham, Mass. IT research firm, makes distinctions can muddy the waters a bit. For IDC, the high-end includes servers that are $250,000 and more, while the midrange is $100,000 to $250,000 and the small servers are less than $100,000.
Under those definitions, the 570 gets split up. Some of the 570s get calculated in the high-end alongside the 590s, and some of them get calculated in the midrange alongside the 550s. In the end, Mauri used these IDC definitions when stating the percentage increases above. But in general, it was interesting to discover what exactly constitutes high-end, and what exactly constitutes midrange, in the industry.
If you’re like me, the recent IBM financial numbers had you wondering. Big Blue gave figures for the “converged System p” servers, which was up 29 percent, and “System i” servers, down 47 percent. In a word, ouch.
But it’s mainly ouch because of the way it makes System i look, and not necessarily because of how the platform actually performed. Let’s take a look at how it all breaks down, according to Ross Mauri, IBM’s general manager of Power Systems.
The “converged System p” line includes all the Power6-based servers that can run AIX and IBM i. When asked why the company doesn’t break that up into subsegments that includes i, Mauri said that they “did do that, internally, but we’re not giving out that information externally. We try to look at what the lead operating system is (for the user) during the purchase, but it’s not something that we can definitively explain.”
Further, the “converged System p” numbers from this quarter — which include the newer System i machines — are compared to the previous year’s System p only numbers. So it makes System p performance look better than reality, although by how much we’re not sure.
Meanwhile, the System i falling-off-the-cliff numbers only include what Mauri called “legacy” System i servers — that is, pre-Power6. Since IBM isn’t actively selling those machines as strongly as the new boxes, and because that number compares to the entire System i line from the year before, the precipitous drop-off isn’t as precipitous as it appears at first.
Mauri would only add that “System p Unix had an outstanding quarter and i had an OK quarter.”
Why IBM decided to report the Power Systems revenue as “converged System p” is a question Mauri couldn’t answer, but it likely has to do with investor expectations. They need points to compare, and a new division called Power Systems probably doesn’t do it for them. But in future years, look for IBM to start reporting numbers under Power Systems, instead of “converged System p.”
The second episode of MoshiMoshi, an interactive Web game sponsored by System i vendor Bytware, is coming to a close this week (July 30, to be exact). We here had written pretty extensively about the first episode, which follows the trials and tribulations of various characters in a fictional corporation as they come across various dilemmas involving IT security.
It’s a good diversion from a typical, mundane day, and there’s a chance you could win prizes such as free consulting or software licenses.
Today RPG is on my mind (perhaps because it was the focus of our Search400.com News newsletter today). And truly, RPG seems to be one of the most contentious issues with the i. For example, Pinnacle Foods recently gave their RPG application a Web revamp when the employees began complaining about using character-based interfaces to access interfaces running RPG. So we want to know, how are you dealing with this complaint at your company? What solutions have you implemented?
Also, I was checking out some of the recent questions over at IT Knowledge Exchange and found a great little discussion about the difference between RPG400 and RPGLE. The comments entered into the answer wiki explain a lot about the differences, and it’s a good reminder that IT Knowledge Exchange is a great resource for the AS/400 community. The AS/400 tag is one of the most popular on the site, and with a high level of participation in both the questioning and answering from some of your fellow AS/400 users it would be a shame for you to miss out.