Suppose you have an application that consumes all the computing power. Now you want to do one more necessary job. Is there any way in Linux to interrupt this application and its state checkpoint so that it can later be resumed from the state that was interrupted?
I am particularly interested in the way in which the application can be stopped and restarted on another machine. Is this also possible?
, ( , , TCP/IP...).
, BLCR .. ( : , ..).
. .
. . Emacs ( Xemacs) unexec.c ( ).
. SBCL ( Common Lisp) . SML/NJ . Squeak ( Smalltalk) .
GCC *.h ( , GCC), .
*.h
. . serialization ( à la JSON, YAML, XML,...). hibernation ( ).
man- man kill
man kill
:
kill -STOP <pid>
kill -CONT <pid>
<pid> - .
<pid>
: Control + Z ( SIGTSTP)
Control + Z
bg/fg,
bg
fg
POPIX . , ; . , , - , , . , - /. , , , .
, , ( ) 128- , ...
On linux, you can send this STOP process signal. Leter you resume it by sending a CONT signal. Please refer to the destruction guide.