Re: [zfs-discuss] weird bug with Seagate 3TB USB3 drive
On Fri, Nov 11, 2011 at 09:55:29PM -0800, Richard Elling wrote: On Nov 10, 2011, at 7:47 PM, David Magda wrote: On Nov 10, 2011, at 18:41, Daniel Carosone wrote: On Tue, Oct 11, 2011 at 08:17:55PM -0400, John D Groenveld wrote: Under both Solaris 10 and Solaris 11x, I receive the evil message: | I/O request is not aligned with 4096 disk sector size. | It is handled through Read Modify Write but the performance is very low. I got similar with 4k sector 'disks' (as a comstar target with blk=4096) when trying to use them to force a pool to ashift=12. The labels are found at the wrong offset when the block numbers change, and maybe the GPT label has issues too. Anyone know if Solaris 11 has better support for detecting the native block size of the underlying storage? Better than ? If the disks advertise 512 bytes, the only way around it is with a whitelist. I would be rather surprised if Oracle sells 4KB sector disks for Solaris systems? Afaik the disks advertise both the physical and logical sector size.. at least on Linux you can see that the disk emulates 512 bytes/sector, but natively it uses 4kB/sector. /sys/block/disk/queue/logical_block_size=512 /sys/block/disk/queue/physical_block_size=4096 The info should be available through IDENTIFY DEVICE (ATA) or READ CAPACITY 16 (SCSI) commands. -- Pasi ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] weird bug with Seagate 3TB USB3 drive
On Nov 12, 2011, at 00:55, Richard Elling wrote: Better than ? If the disks advertise 512 bytes, the only way around it is with a whitelist. I would be rather surprised if Oracle sells 4KB sector disks for Solaris systems… Solaris 10. OpenSolaris. But would it be surprising to use SANs with Solaris? Or perhaps run Solaris under some kind of virtualized environment where the virtual disk has a particular block size? Or maybe SSDs, which tend to read/write/delete in certain block sizes? In these situations simply assuming 512 may slow things down. And if Solaris 11 is going to be around for a decade or so, I'd hazard to guess that 512B sector disks will become less and less prevalent as time goes on. Might as well enable the functionality now, when 4K is rarer, so you have more time to test and tunes things out—rather than later when you can potentially be left scrambling. As Pasi Kärkkäinen mentions, there's not much you can do if the disks lies (just as has been seen with disks that lie about flushing the cache). This is mostly a temporary kludge for legacy's sake. More and more disks will be truthful as times goes on. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] weird bug with Seagate 3TB USB3 drive
On Sat, Nov 12, 2011 at 08:15:31AM -0500, David Magda wrote: On Nov 12, 2011, at 00:55, Richard Elling wrote: Better than ? If the disks advertise 512 bytes, the only way around it is with a whitelist. I would be rather surprised if Oracle sells 4KB sector disks for Solaris systems? Solaris 10. OpenSolaris. But would it be surprising to use SANs with Solaris? Or perhaps run Solaris under some kind of virtualized environment where the virtual disk has a particular block size? Or maybe SSDs, which tend to read/write/delete in certain block sizes? In these situations simply assuming 512 may slow things down. And if Solaris 11 is going to be around for a decade or so, I'd hazard to guess that 512B sector disks will become less and less prevalent as time goes on. Might as well enable the functionality now, when 4K is rarer, so you have more time to test and tunes things out?rather than later when you can potentially be left scrambling. As Pasi Kärkkäinen mentions, there's not much you can do if the disks lies (just as has been seen with disks that lie about flushing the cache). This is mostly a temporary kludge for legacy's sake. More and more disks will be truthful as times goes on. Most 4kB/sector disks already today properly report both the physical (4kB) and logical (512b) sector sizes. It sounds like *solaris is only checking the logical (512b) sector size, not the physical (4kB) sector size.. -- Pasi ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] weird bug with Seagate 3TB USB3 drive
On Nov 12, 2011, at 8:31 AM, Pasi Kärkkäinen wrote: On Sat, Nov 12, 2011 at 08:15:31AM -0500, David Magda wrote: On Nov 12, 2011, at 00:55, Richard Elling wrote: Better than ? If the disks advertise 512 bytes, the only way around it is with a whitelist. I would be rather surprised if Oracle sells 4KB sector disks for Solaris systems? Solaris 10. OpenSolaris. But would it be surprising to use SANs with Solaris? Or perhaps run Solaris under some kind of virtualized environment where the virtual disk has a particular block size? Or maybe SSDs, which tend to read/write/delete in certain block sizes? In these situations simply assuming 512 may slow things down. And if Solaris 11 is going to be around for a decade or so, I'd hazard to guess that 512B sector disks will become less and less prevalent as time goes on. Might as well enable the functionality now, when 4K is rarer, so you have more time to test and tunes things out?rather than later when you can potentially be left scrambling. As Pasi Kärkkäinen mentions, there's not much you can do if the disks lies (just as has been seen with disks that lie about flushing the cache). This is mostly a temporary kludge for legacy's sake. More and more disks will be truthful as times goes on. Most 4kB/sector disks already today properly report both the physical (4kB) and logical (512b) sector sizes. It sounds like *solaris is only checking the logical (512b) sector size, not the physical (4kB) sector size.. ZFS uses the physical block size. http://src.illumos.org/source/xref/illumos-gate/usr/src/uts/common/fs/zfs/vdev_disk.c#294 -- richard -- ZFS and performance consulting http://www.RichardElling.com LISA '11, Boston, MA, December 4-9 ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Hare receiving snapshots become slower?
On 09/30/11 08:12 AM, Ian Collins wrote: On 09/30/11 08:03 AM, Bob Friesenhahn wrote: On Fri, 30 Sep 2011, Ian Collins wrote: Slowing down replication is not a good move! Do you prefer pool corruption? ;-) Probably they fixed a dire bug and this is the cost of the fix. Could be. I think I'll raise a support case to find out why. This is making it difficult for me to meet a replication guarantee. It turns out this was a problem with e1000g interfaces. When we swapped over to an igb port, the problem went away. -- Ian. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
