Hi all I have heard/read about systems from Sun and NetApp using SDs for caching. Sun uses different SDs for read and write caching (one being faster for reads and the other for writes). The storage solution from Sun is based on Solaris with ZFS. I know too little about filesystems design to know exactly how they do this, but can someone think of a way this could be implemented in btrfs? I would think such a cache level could speed up things rather well. Best regards roy -- Roy Sigurd Karlsbakk (+47) 97542685 roy@karlsbakk.net http://blogg.karlsbakk.net/ -- I all pedagogikk er det essensielt at pensum presenteres intelligibelt. Det er et elementært imperativ for alle pedagoger å unngå eksessiv anvendelse av idiomer med fremmed opprinnelse. I de fleste tilfeller eksisterer adekvate og relevante synonymer på norsk. -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 18. juni. 2009, at 15.56, Roy Sigurd Karlsbakk wrote:> Hi all > > I have heard/read about systems from Sun and NetApp using SDs for > caching. Sun uses different SDs for read and write caching (one > being faster for reads and the other for writes). The storage > solution from Sun is based on Solaris with ZFS. > > I know too little about filesystems design to know exactly how they > do this, but can someone think of a way this could be implemented in > btrfs? I would think such a cache level could speed up things rather > well.erm. Obviously, I meant Solid State Drives, SSDs. Sorry about this. roy -- Roy Sigurd Karlsbakk (+47) 97542685 roy@karlsbakk.net http://blogg.karlsbakk.net/ -- I all pedagogikk er det essensielt at pensum presenteres intelligibelt. Det er et elementært imperativ for alle pedagoger å unngå eksessiv anvendelse av idiomer med fremmed opprinnelse. I de fleste tilfeller eksisterer adekvate og relevante synonymer på norsk. -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Jun 18, 2009 at 03:56:23PM +0200, Roy Sigurd Karlsbakk wrote:> I have heard/read about systems from Sun and NetApp using SDs for > caching. Sun uses different SDs for read and write caching (one being > faster for reads and the other for writes). The storage solution from > Sun is based on Solaris with ZFS. > > I know too little about filesystems design to know exactly how they do > this, but can someone think of a way this could be implemented in btrfs?Linux already has second-level¹ cache layer, it called FS-Cache. For now, the sole user is NFS. btrfs could take advantage of FS-Cache also. ¹ first-level is page cache in RAM -- Tomasz Torcz "Never underestimate the bandwidth of a station xmpp: zdzichubg@chrome.pl wagon filled with backup tapes." -- Jim Gray -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 18. juni. 2009, at 17.55, Tomasz Torcz wrote:> On Thu, Jun 18, 2009 at 03:56:23PM +0200, Roy Sigurd Karlsbakk wrote: >> I have heard/read about systems from Sun and NetApp using SDs for >> caching. Sun uses different SDs for read and write caching (one being >> faster for reads and the other for writes). The storage solution from >> Sun is based on Solaris with ZFS. >> >> I know too little about filesystems design to know exactly how they >> do >> this, but can someone think of a way this could be implemented in >> btrfs? > > Linux already has second-level¹ cache layer, it called FS-Cache. > For now, > the sole user is NFS. btrfs could take advantage of FS-Cache also.I googled around a bit for this, but couldn''t find anything useful. Where can I find anything about this second (or fifth) level caching? roy -- Roy Sigurd Karlsbakk (+47) 97542685 roy@karlsbakk.net http://blogg.karlsbakk.net/ -- I all pedagogikk er det essensielt at pensum presenteres intelligibelt. Det er et elementært imperativ for alle pedagoger å unngå eksessiv anvendelse av idiomer med fremmed opprinnelse. I de fleste tilfeller eksisterer adekvate og relevante synonymer på norsk. -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Jun 18, 2009 at 07:25:46PM +0200, Roy Sigurd Karlsbakk wrote:> On 18. juni. 2009, at 17.55, Tomasz Torcz wrote: > > > On Thu, Jun 18, 2009 at 03:56:23PM +0200, Roy Sigurd Karlsbakk wrote: > >> I have heard/read about systems from Sun and NetApp using SDs for > >> caching. Sun uses different SDs for read and write caching (one being > >> faster for reads and the other for writes). The storage solution from > >> Sun is based on Solaris with ZFS. > >> > >> I know too little about filesystems design to know exactly how they > >> do > >> this, but can someone think of a way this could be implemented in > >> btrfs? > > > > Linux already has second-level? cache layer, it called FS-Cache. > > For now, > > the sole user is NFS. btrfs could take advantage of FS-Cache also. > > > I googled around a bit for this, but couldn''t find anything useful. > Where can I find anything about this second (or fifth) level caching?Will this help? http://www.linux-mag.com/cache/7378/1.html -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Tomasz Torcz a écrit :> Linux already has second-level¹ cache layer, it called FS-Cache. For now, > the sole user is NFS. btrfs could take advantage of FS-Cache also. > > ¹ first-level is page cache in RAM >FS-Cache is a good thing to speedup your network filesystem, but to be efficient localy, IMHO, the second-level cache shoud be at the block level. So it should go directly in btrfs or in something like the device mapper ... Sébastien Wacquiez -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html