Re: question on concurrent code upgrades in AIX&ESS environments

I worked at a large financial services firm that mandated downtime for all microcode changes, regardless of Satan's empty promises. The one exception to this rule was a fix to a serious problem. In this case, there are many Satan's!

Generally, for a production shop, I'd set a service level for applying maintenance like this and then schedule an outage, complete with a backout plan. The service level, like system software maintenance, is stated as something like X months behind current where x is determined by a sine wave regression analysis between your intestinal fortitude and your negotiating skill.

Especially for a box like the Shark, all the really bad errors have been found and fixed by now so why fix something that ain't really broke.


"Miller, Dave (I.S.)" <Dave.Miller@xxxxxxx> wrote:
.EmailQuote { margin-left: 1pt; padding-left: 4pt; border-left: #800000 2px solid; } We have had a number of challenges for the past 3-4 years doing concurrent firmware/code upgrades to the ESS (we use 2105-800). The hardware is sold as concurrent code update capable and of course (rightfully so) the powers that be feel we should not need downtime to accomplish them?

We have had at least 3 service affecting episodes with them (downtime caused by LVM errors in AIX) that was directly relate to the ESS, or in one case SDD.

- Am I the only one that has this kind of ?luck??
- does your site allow concurrent updates, or do you take downtime?

I would truly appreciate any and all feedback. We really need to do a ?sanity check? for this process and re-think it. thanks.

Note: if you are reluctant to clutter up the list or if folks feel its off-topic please e-mail me directly. Dave.miller@xxxxxxxx





---------------------------------
David Milller
Baystate Health System
Springfield, MA
dave.miller@xxxxxxx

-------------------------------------------------------------------------------------------------------------------------------------------------
Fools ignore complexity. Pragmatists suffer it. Experts avoid it. Geniuses remove it (Alan J. Perlis)
-------------------------------------------------------------------------------------------------------------------------------------------------





Edward Long

Relevant Pages

  • Re: Great SWT Program
    ... because you didn't fix them. ... Hardware problems cause downtime. ... Hardware that magically continues to work after it fails, ...
    (comp.lang.java.programmer)
  • Re: Great SWT Program
    ... because you didn't fix them. ... HARDWARE PROBLEMS CAUSE DOWNTIME, ... powered by emacs ...
    (comp.lang.java.programmer)