[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Timeline for potato

On Fri, 31 Dec 1999, Michael Stone wrote:
>On Fri, Dec 31, 1999 at 05:03:14PM +1100, Russell Coker wrote:
>> That's debatable.  However one thing that's not disputed is that 2.2.14 is
>> due out really soon (but not before Jan 2 I expect).  2.2.14 will just be
>> more of the same as 2.2.13, it apparently solves some reliability issues and
>> will hopefully fix the software RAID problems that make RAID-1 and RAID-5
>> unsuitable for ReiserFS.
>AFAIK, there's no intention of including new-raid in any 2.2 kernel.
>(There's still argument about whether it will go into 2.4. :( Backward
>compatability sucks, and seems bizzare in this case.)

My previous message wasn't as clear as it should have been.

I was not referring to the new RAID (aka Raidtools2), but to the patch to the
old RAID which will stop it from writing out dirty blocks before they are
ready to be written.
Currently if you have a RAID-5 setup and you have a dirty block that is not
scheduled to be written (ordering issue related to a journalling file system)
then a write to another block in the same stripe can cause the dirty block to
be written.
I believe that something similar can occur with RAID-1 but it's significantly
less likely.

The ultimate result is that some innovations that would truly benefit
consumers never occur for the sole reason that they do not coincide with
Microsoft's self-interest.
-- Judge Thomas Penfield Jackson, U.S. District Judge

Reply to: