Messages in this thread | | | From | <> | Date | Mon, 6 Dec 1999 08:15:02 -0500 (EST) | Subject | Process "pickling" |
| |
==> Regarding Process "pickling"; "Curtis M. Brune" <curtis_brune@workspot.com> adds:
curtis_brune> Hey, Is there any movement in the kernel world to "pickle" curtis_brune> process, i.e. completely move a process and all associated curtis_brune> kernel resources to disc. At a later time the process could curtis_brune> be "brought back to life" on a DIFFERENT machine. The idea curtis_brune> would be to move processes around between a cluster of curtis_brune> machines, possiblly storing a process for months or years curtis_brune> before waking it up again.
curtis_brune> Has this been a discussed before? Is it even possible? curtis_brune> Seems very useful.
If you mean checkpoint/restart and/or process migration, then yes. There are patches you can obtain to do this. See bproc:
http://beowulf.gsfc.nasa.gov/software/bproc.html
There is a less portable set of patches named epckpt found here:
http://www.cs.rochester.edu/~edpin/epckpt/
Currently I am working on a project to perform transparent process migration. It was originally based on epckpt, but then we moved to bproc. There is no useful code as of now. If you would like to track our progress, see:
http://segfault.dhs.org/pants/
Regards,
Jeff
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |