>I can understand the reason for those IEFUSI warnings. Java! Well, the warnings for IEFUSI were there long before Java came along. And it was only the precursor for things to come.
>Also those of us who compile using the C/C++ compiler in USS need a lot of >memory because the compiler is a memory hog when optimizing. It's no >different when compiling using batch. Not to mention a cpu hog. >Please name and shame the application! Sounds like a forker. Can I say the good old 'check the archives'? I have complained about it at length. >It's also not uncommon to see Java applications using thread pools with 50 >times more threads then cpus. As soon as notifyAll is called the thundering >herd comes a running. Most of these problems are due to poor programming >practice, not necessarily the environment. My point of contention is that most of the 'programmers' (That's why I called that 'clicking') don't care that their code is poor. My neighbour - a nice young man of 25, just finished his IT-education, and he is sharp! - stated the mind set of those I call 'clickers': "If it works on my PC, I don't care if it has a performance problem in production. Someone else in the project hierarchy has to fix it." (like the architect for the project or the customer). With this attitude, about 99% of the 'ported' code is really bad for the environment it is supposed to run in productively. And the 1% that isn't so bad has a lot of customer blood attached to it. >I guess IBM are building hardware to make it >perform (z10) just like they did back in the day for DB2. Don't know about DB2, but performance problems caused by extremely poor programming in a zLinux environment were dealt with by doing what is done in the mickey mouse world - throw more hardware at it (in our case more IFLs). That one single application cannot function with less than 8 IFLs. To compare - our 'holy cow' production system has only 7 general cps, and it runs IMS, DB2 and tons of batch without them crashing, even when we execute 1400 IMS transactions per second for at least a minute (with response times *a lot* less than 50ms). Guess I should shut up now. Barbara ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
