I have not said that it will not be good for some users. I have said
that it does nothing to solve problems that we have had in the past and
will most likely have again. Without the other piece, dynamic removal of
storage, the only solution we see for our problems, when we become
constrained, is to buy more iron. If the dynamic removal were included,
it could delay upgrades and save money. Yes, being able to add new
storage will be nice. When we have a z10 with z/VM 5.4, the memory
upgrades may become easier; however, there is still the question of the
storage that is getting moldy waiting for its LPAR to be activated.
Storage that could be used if there were some way to release it without
the LPAR activation outage. Having the release, that or having LPAR
storage virtualized, would be the best of all possible worlds (until
someone sees another pie floating in the stratosphere).
Regards,
Richard Schuh
________________________________
From: The IBM z/VM Operating System
[mailto:[EMAIL PROTECTED] On Behalf Of James Vincent
Sent: Tuesday, August 05, 2008 11:29 AM
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: Some IBM Announcements for z/OS, z/VM, z/VSE (Aug
5, 2008)
Richard, there are good number of folks that -will- find this
useful to an unbelievable level. Imagine having LPARs loaded with Linux
guests - hundreds of them that cross all lines of business at a company.
Any outage on any of those LPARs takes over six to eight hours and will
only be allowed on Sundays starting at midnight, because everyone and
their little sister has to be involved to check, fix and diagnose issues
on their applications after things start back up (or don't).
Now imagine that company is growing the Linux LPARs because more
and more work is going there, and every quarter they need to add 16-32G
on each LPAR. Taking outages like that are painful. Very painful. The
'technical' part of the outage is the easy part; dealing with
apps/servers that are not behaving well or application support that just
don't understand how things work is the hard part.
That company can now save the outage time, the time required for
planning the outages, probable outages later due to miss-configured
apps/servers, etc etc.
With z/VM 5.4, you issue a command and the memory is there
(assuming you have done good planning and set the reserve memory when
you installed z/VM 5.4!)
I do completely understand what you mean in your specific case
though. You have spare memory on your machine you "could" use, but
would have to give it back when those other LPARs need to run. Maybe
there are others like you that require the "release the memory" and
would consider entering a requirement to IBM for it. I am sure that
concept was probably discussed at some point in the design - it had to
be. Any z/VM'er worth their salt will ask "if I can turn it on, can I
turn it off? Can I have different colors, hot coffee and an extra week
off work with it too?"
Just getting the ability to add real memory dynamically is a
huge boon and lays the foundation for possibly more feature/function
later on.
Jim Vincent
On Tue, Aug 5, 2008 at 12:38 PM, Schuh, Richard
<[EMAIL PROTECTED]> wrote:
...snip... The new capability is
absolutely useless to us. Nice boilerplate to show in
sales pitches to
management, perhaps, but no practical value here. Our VM
is as close to
24 X 365.25 as we can make it. Taking it down to
reconfigure the storage
is unacceptable.
Regards,
Richard Schuh
