(not that I want to date myself) but in the early eighties we were all rooting for Microsoft to topple IBM, or at least give it a good kick in the pants. IBM was a big arrogant bloated lumbering beast that did not tread lightly on their competition (sound familiar?). They were definately taught a lesson.
Now, here I am rooting for IBM to kick Microsoft in the pants, and but good. You go IBM!!!
First off, lets exclude the "program the suspension into your software" ideas. Not that they are not valid (they are extremely valid), but lets say that you either do not have access to the code or the integration of suspend/recover in the code is prohibitive because of complexity, time to test, resources, etc. Also, lets exclude the "whole system shutdown" approach, which is covered by other software.
What you are really looking for here is two fold: You want to be able to suspend/recover more that one program associated with a particular task (lets say a program with its associated database tasks) AND you need to coordinate the suspension/recovery.
So what you need is a coordinated core system that cores out the processes you tell it to (along with child processes) at a single specific time. The coordinated core system would then be able to restart the processes at a later date.
Wow! That would be cool. How many times have I wanted to freeze a set of processes, reboot a box, and then start them up where they left off. It would be a great debugging tool!
Slashdot Mirror
(not that I want to date myself) but in the early eighties we were all rooting for Microsoft to topple IBM, or at least give it a good kick in the pants. IBM was a big arrogant bloated lumbering beast that did not tread lightly on their competition (sound familiar?). They were definately taught a lesson. Now, here I am rooting for IBM to kick Microsoft in the pants, and but good. You go IBM!!!
What you are really looking for here is two fold: You want to be able to suspend/recover more that one program associated with a particular task (lets say a program with its associated database tasks) AND you need to coordinate the suspension/recovery.
So what you need is a coordinated core system that cores out the processes you tell it to (along with child processes) at a single specific time. The coordinated core system would then be able to restart the processes at a later date.
Wow! That would be cool. How many times have I wanted to freeze a set of processes, reboot a box, and then start them up where they left off. It would be a great debugging tool!
OK, feeling a little giddy.... deep breaths....