On 11/14/05, Eric Lowe <[EMAIL PROTECTED]> wrote:
Well cheap ram may rule out its use in big projects and corporations, there are other limits about ram used in a system on the lowend. Many people still use Ultra 5/10 that have a hardware limit of 1GB ram limit, even modern systems will see benefits because of hardware ram limits even the u20 has a maximum ram limit of just 4GB, if compressed cache would give an user better performance, because of it. I would see it being quite popular.
.
If this is the case, compressed cache may become a mute point since last i heard zfs allowed compression of its data and it kept the data compressed as well when it was cached.
James Dickens
uadmin.blogspot.com
On Mon, Nov 14, 2005 at 10:28:38AM -0800, Nitin Gupta wrote:
| Hi,
| I've been working on porting the 'compressed cache' feature
| (http://linuxcompressed.sourceforge.net/ - feature explained in first few lines)
| to linux 2.6 kernel.
| I'm wondering why this project is dead even when it showed great
| performance improvement when system is under memory pressure.
While it sounds interesting from an academic point of view, I wonder
how relevant this work (and any work related to paging) is when memory
can be purchased for under $200 a gigabyte. This is a different perspective
than Linux which is in common use on older machines and smaller machines
(even embedded systems) with small memories -- Solaris typically runs on
servers requiring a specified response time. Solaris users usually
consider it a *failure mode* when a system starts paging since its
response time becomes non-linearly poor and unpredictable!
Well cheap ram may rule out its use in big projects and corporations, there are other limits about ram used in a system on the lowend. Many people still use Ultra 5/10 that have a hardware limit of 1GB ram limit, even modern systems will see benefits because of hardware ram limits even the u20 has a maximum ram limit of just 4GB, if compressed cache would give an user better performance, because of it. I would see it being quite popular.
.
Hence, I don't think this or any paging related projects will generate a
lot of interest in Solaris.
| Are there any serious drawbacks to this?
| Do you think it will be of any use if ported to solaris kernel?
As far as I can see, this is nearly impossible to port to Solaris, which
unlike Linux allows direct mapping of pages in the page cache. Keeping
compressed data pages around in memory would also require having a better
idea of candidate pages to pageout than we can achieve on SPARC CPUs
since these CPUs do not have hardware support for ref/mod bits (making
paging candidate determination very costly). ZFS is very fast (being copy-
on-write it does sequential I/Os whereas swapfs does random disk access),
and it supports compression. ZFS is already slated to become the new
replacement to swapfs in the near future.
If this is the case, compressed cache may become a mute point since last i heard zfs allowed compression of its data and it kept the data compressed as well when it was cached.
James Dickens
uadmin.blogspot.com
--
Eric Lowe Solaris Kernel Development Austin, Texas
Sun Microsystems. We make the net work. x64155/+1(512)401-1155
_______________________________________________
perf-discuss mailing list
perf-discuss@opensolaris.org
_______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
