Re: [RFC v13][PATCH 00/14] Kernel based checkpoint/restart

Oren Laadan wrote:

Hi,

Just got back from 3 weeks with practically no internet, and I see
that I missed a big party !

Trying to catch up with what's been said so far --

[...]

quoted

quoted

quoted

- Will any of this involve non-trivial serialisation of kernel
  objects?  If so, that's getting into the
  unacceptably-expensive-to-maintain space, I suspect.

We have some structures that are certainly tied to the kernel-internal
ones.  However, we are certainly *not* simply writing kernel structures
to userspace.  We could do that with /dev/mem.  We are carefully pulling
out the minimal bits of information from the kernel structures that we
*need* to recreate the function of the structure at restart.  There is a
maintenance burden here but, so far, that burden is almost entirely in
checkpoint/*.c.  We intend to test this functionality thoroughly to
ensure that we don't regress once we have integrated it.

I guess my question can be approximately simplified to: "will it end up
looking like openvz"?  (I don't believe that we know of any other way
of implementing this?)

Because if it does then that's a concern, because my assessment when I
looked at that code (a number of years ago) was that having code of
that nature in mainline would be pretty costly to us, and rather
unwelcome.

I originally implemented c/r for linux as as kernel module, without
requiring any changes from the kernel. (Doing the namespaces as a kernel
module was much harder). For more details, see:
	https://www.ncl.cs.columbia.edu/research/migrate

oops... I meant the following link:
	http://www.ncl.cs.columbia.edu/research/migration/

see, for example, the papers from DejaView (SOSP 07) and Zap (USENIX 07).

Oren.


--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>