Well, the simplest case is service office and contract containing the
fee for unexpected downtime. The money has to be paid.
However in other cases there are some business estimations. Sometimes
quite reasonable, in many cases not really. However it is also
unreasonable to judge the case without deep knowledge of details.
Last, but not least: huge numbers are good to keep discipline. ;-)
--
Radoslaw Skorupka
Lodz, Poland
W dniu 04.03.2025 o 16:06, Dick Williams pisze:
Worked at GM subsidiary Packard Electric in the mid 80’s. Back then if the
mainframe was down, it costs Packard 200k per hour. And it was absolutely true.
Thousands of people getting paid top manufacturing dollars twiddling their
thumbs tends to do that.
Sent from Yahoo Mail for iPhone
On Tuesday, March 4, 2025, 8:54 AM, Phil Smith III<li...@akphs.com> wrote:
Of course. But that's the point: without other data, the only metric can be
revenue. And there are tons of ways for it to be wrong in either direction, so
claimed costs are (almost?) always fantasy.
-----Original Message-----
From: IBM Mainframe Discussion List<IBM-MAIN@LISTSERV.UA.EDU> On Behalf Of
Radoslaw Skorupka
Sent: Monday, March 3, 2025 9:20 PM
To:IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: Cost of an outage
W dniu 03.03.2025 o 15:27, Phil Smith III pisze:
[...]
Similarly, Allan Staller wrote, in part:
At a relatively small MF shop I used to work at, the cost of downtime
was pegged at 100K/Hour.
Again, could be true (though that seems to make the company's revenue $876M,
which doesn't sound like a small shop) but where did that number come from?
Good calculation, wrong assumption.
Indeed, the cost of outage need not to be the revenue made during online
processing.
In other words the company can earn much less than they loss when unexpected
outage occurs.
And of course the cost of outage is provided in $$$/s while the real cost per
time is rarely linear.
There are cases where <1 minute outage would not cause any measurable costs,
while 5 minutes causes big problems, same problems for 10 minutes, but 30m and
more means even worse problems, but there is no difference between 2 and 4 hours,
etc. That include reputation also, which is hardly measurable.
--
Radoslaw Skorupka
Lodz, Poland
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
