Messages in this thread | | | Date | Tue, 22 Jun 1999 16:13:47 +0200 (CEST) | From | Ingo Molnar <> | Subject | Re: New buffer/cache small report and questions |
| |
On Tue, 22 Jun 1999, Ricardo Galli Granada wrote:
> Congratulations, the results look very good.
great.
> My question is, are the changes going to be back ported to 2.2.x?
i dont think so. But the next stable release is not far away. The development branch will always have 'cool stuff, wish we had that in the stable branch too', hopefully in the future stable branches will follow up each other with higher frequency.
> My second question is, can a 4096 bytes block size improve results even > more?
definitely! Many of the improvements show up in a magnified way on 4k block filesystems. It would be very interesting to see the same measurement done on a 4k-blocksize testpartition.
as a rule of thumb, ext2 filesystems should always be created with 4k blocksize, unless there will be lots of small files on the filesystem. (typically not the case with most database systems)
-- mingo
ps. be careful with that production system though, even pre2-2.3.8 has known, serious, not yet fixed bugs which might corrupt files.
- 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/
| |