>Hmmm... "In some situations, your operations may be so critical that
>human analysis and intervention may not be fast enough, and the system
>must have the ability to act quickly and decisively."

yes, I have been laughing real hard about that one! Especially in light of z/OS 
still being so dumb as to schedule more than 10 dumps in less than a second 
for an 878-10, effectively tying up the system and drastically increasing 
paging. Talk about absolutely intelligent recovery routines!

Did I mention scheduled dumps cannot be suppressed via slip? And apparently 
*someone* made the decision to *schedule* dumps again (this is new in z/OS 
1.10, 1.8 didn't have this, and we always had TSO/Xpediter users who are 
prone to this 878.)
Did I mention that DAE is useless in this regard (too many dump requests 
before the first one has its symptom string written?)

Once we migrate to 1.12, we'll probably look for a lot of ways to turn off 
this 'predictive failure analysis' feature. Might be as bad as (some of the) 
health checker checks that are not useful in the real world of smaller 
installations.

I also laughed real hard about this:
"The z/OS Management Facility (z/OSMF, 5655-S28) is the new face for z/OS
and it provides support for a modern, Web-browser-based management console
for z/OS. Automated tasks and wizards can guide users through tasks and help
provide simplified operations. In z/OSMF V1.11, for example, tasks taking up to
20 minutes, such as collecting and packaging dump data, can now take as 
little as 30 seconds."

Yes, let's take dumps via the wizards! That will definitely improve software 
support. Also, given that IBMs own installations are not allowed to directly 
ftp 
to IBM (and almost everyone has to fight the firewall fractions), I wonder 
how "collecting and packaging dump data" will only take 30seconds using a 
wizard. (Especially when encrypting the dump alone takes several cpu minutes 
already.)  Does IBM now also have a 'wizard' to actually *read* the dump?!?

I did notice, though, that "RSM and dump processing will be designed to 
improve capture performance for large amounts of data during SVC dump 
processing." This will be a good thing! 

Regards, Barbara Nitz

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to