Fwd: 1 minutes: best downtime story
- From: Nigel Thomas <nigel.cl.thomas@xxxxxxxxxxxxxx>
- To: Oracle-L Freelists <oracle-l@xxxxxxxxxxxxx>
- Date: Fri, 15 Mar 2013 17:35:02 +0000
Late 80s, Oracle UK launched Business Applications (later superseded by the
US-developed apps we now know as "Oracle Apps"). On Feb 29 1988 every
single customer's application stopped working. Turned out that the
SQL*Forms 2 user exit that verified the date went something like this:
1) Get the dayNumber and month
2) Check 1 <= dayNumber < maxDays[month] else fail
3) If it is a leap year and Feb, check that 1 <= dayNumber <= 29 else fail
Sadly the exception raised at step 3 prevented the "oh, but it's OK after
all" step 4. And of course, this was the first leap year ever for this bit
of code, which was only there to avoid the typical SELECT ... FROM DUAL
post change trigger.
Workarounds: go home, come back tomorrow - or just reset your system clock
one day forward or back.
Regards
On 14 March 2013 21:07, Jeremy Schneider <jeremy.schneider@xxxxxxxxxxxxxx>wrote:
> Hey all -
>
> I'm writing a paper about top causes of downtime. As one component
> of research, I'd like to get some input from you!
>
> One minute, two sentences. First sentence: describe what went down.
> Second sentence: describe why. (I have to categorize all of these.)
> Everyone should have at least one downtime story so I'm hoping for a
> lot of feedback!
>
> Answer about any technology - database, operating system, etc.
>
> Thanks!
>
> -Jeremy
>
>
> --
> Jeremy Schneider
> Pythian Consulting Group
> Chicago
>
> +1 312-725-9249
> http://www.pythian.com
>
> --
> //www.freelists.org/webpage/oracle-l
>
>
>
--
//www.freelists.org/webpage/oracle-l
Other related posts:
- » Fwd: 1 minutes: best downtime story - Nigel Thomas
