Messages in this thread | | | Date | Wed, 01 Oct 1997 12:48:46 -0400 | From | David Schwartz <> | Subject | Re: Process Migration on Linux - Impossible? |
| |
At 11:05 AM 10/1/97 -0500, Tall cool one wrote: >Victor Yodaiken <yodaiken@chelm.cs.nmt.edu> writes: >> On Tue, Sep 30, 1997 at 10:06:25PM -0500, Tall cool one wrote: >> > Personally, I don't have a real need for process migration, but I'd like >> > to see it because of what process migration would allow. The ability of >> >> So you want fault tolerance, not process migration. Don't confuse the >> technique with the result. > > That's an odd thing to say, isn't it? Re-read my first sentence again, >carefully this time. If one can throw a process on another machine at will, >it becomes trivial to checkpoint, yes? Fault tolerance for free.
Nope. Not so. You can't migrate a TCP/IP connection between machines even if you migrate the process that owns it. Fault tolerance and process migration are 100% separate issues.
> Okay, I'd like to see proccess migration, process checkpointing, and fault >tolerance (especially fault tolerance). Whatever makes you happy. > > Typically I wouldn't need to migrate network applications, although it >could conceivably be possible with some sort of proxying. Not that I'd want >to implement it. =) Or perhaps the machines share the same IP's but have an >running agreement on what ports on which machines the IP's go to. Probably >could be done with local firewall rules. Need a userland daemon to keep the >"port-map" in sync though.
Network I/O is not the only problem. File I/O is just as important and just as serious.
DS
| |