Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-22 Thread Saso Kiselkov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I didn't mean to imply that I use it for my media storage, just that I occasionally encounter situations when it could be useful. BR, - -- Saso On 07/22/2010 11:23 AM, Roy Sigurd Karlsbakk wrote: > - Original Message - >> I do encounter situa

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-22 Thread Roy Sigurd Karlsbakk
- Original Message - > I do encounter situations when I (or somebody from my family) > accidentally create multiple copies of photo albums. :-) I wouldn't recommend using dedup on this system. Dedup requires lots of RAM or L2ARC, and I don't think it is suitable for your needs. You may wa

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Erik Trimble
On 7/21/2010 7:47 AM, Eugen Leitl wrote: How badly would a dual-core 1.6 GHz Atom with 4 GBytes RAM be underpowered for serving 4-6 SATA drives? What kind of transfer speed (GBit Ethernet, Intel NICs) can I expect with raidz2 or raidz3? Thanks. For light usage (I'm assuming you want to us

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Garrett D'Amore
On Wed, 2010-07-21 at 17:12 +0200, Saso Kiselkov wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > If you plan on using it as a storage server for multimedia data > (movies), don't even bother considering compression, as most media files > already come heavily compressed. Dedup might st

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Saso Kiselkov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I do encounter situations when I (or somebody from my family) accidentally create multiple copies of photo albums. :-) - -- Saso On 07/21/2010 05:20 PM, Edward Ned Harvey wrote: >> From: zfs-discuss-boun...@opensolaris.org [mailto:zfs-discuss- >> bou

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Edward Ned Harvey
> From: zfs-discuss-boun...@opensolaris.org [mailto:zfs-discuss- > boun...@opensolaris.org] On Behalf Of Saso Kiselkov > > If you plan on using it as a storage server for multimedia data > (movies), don't even bother considering compression, as most media > files > already come heavily compressed.

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Darren J Moffat
On 21/07/2010 16:12, Saso Kiselkov wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 If you plan on using it as a storage server for multimedia data (movies), don't even bother considering compression, as most media files already come heavily compressed. Dedup might still come in handy, thoug

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Saso Kiselkov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 If you plan on using it as a storage server for multimedia data (movies), don't even bother considering compression, as most media files already come heavily compressed. Dedup might still come in handy, though. - -- Saso On 07/21/2010 05:03 PM, Eugen

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Eugen Leitl
On Wed, Jul 21, 2010 at 04:56:26PM +0200, Roy Sigurd Karlsbakk wrote: > It'll probably be ok. If you use lzjb compresion, it'll probably suffice as > well. Give it gzip-9 compression, and you might have a cpu bottleneck, but > then, for most use, that config will probably do. What sort of traffi

Re: [zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Roy Sigurd Karlsbakk
- Original Message - > How badly would a dual-core 1.6 GHz Atom with 4 GBytes RAM > be underpowered for serving 4-6 SATA drives? What kind of > transfer speed (GBit Ethernet, Intel NICs) can I expect with > raidz2 or raidz3? It'll probably be ok. If you use lzjb compresion, it'll probably

[zfs-discuss] CPU requirements for zfs performance

2010-07-21 Thread Eugen Leitl
How badly would a dual-core 1.6 GHz Atom with 4 GBytes RAM be underpowered for serving 4-6 SATA drives? What kind of transfer speed (GBit Ethernet, Intel NICs) can I expect with raidz2 or raidz3? Thanks. -- Eugen* Leitl http://leitl.org";>leitl http://leitl.org