Re: [zfs-discuss] ZFS/WAFL lawsuit
Joerg Schilling wrote: David Hopwood [EMAIL PROTECTED] wrote: Al Hopper wrote: So back to patent portfolios: yes there will be (public and private) posturing; yes there will be negotiations; and, ultimately, there will be a resolution. All of this won't affect ZFS or anyone running ZFS. It matters a great deal what the resolution is. The best outcome, for everyone wanting to use any COW and/or always-consistent-on-disk filesystem (including btrfs and others), would be for the invalidation part of NetApp's lawsuit to succeed, and the infringement part to fail. A cross-licensing deal or other out-of-court settlement would be much less helpful. Invalidating COW filesystem patents would of course be the best. Unfortunately those lawsuits are usually not handled in the open and in order to understand everything you would need to know about the background interests of both parties. IANAL, but I was under the impression that it was possible to file an amicus brief or amicus curiae, which in this case would detail known prior art -- whether or not it is prior art that benefits either party in the case. http://en.wikipedia.org/wiki/Amicus_curiae The EFF does this quite often, and it has a patent busting project which is described at http://www.eff.org/patent/. (The EFF wanted list does not currently include the WAFL and ZFS patents, but it's only been a few days since the NetApp suit was announced.) It is true that, as the WP article says, The decision whether to admit the information lies with the court. But surely, even in East Texas, it would be difficult to completely dismiss prior art on COW and always-consistent- on-disk filesystems, no matter how submitted. Or am I being too naive? -- David Hopwood [EMAIL PROTECTED] ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS/WAFL lawsuit
George William Herbert wrote: http://blogs.netapp.com/dave/2007/09/netapp-sues-sun.html Curiously, I posted to the blog comments last night discussing some of the prior art, going back to some of the disks could do this too discussions by early tree structured binary data structures inventions, mentioning other copy-on-write structure ideas floating around in the late 80s and early 90s, etc. And it went away. There's only a very generic fake post with my name on it there this morning. Either they have a bug, or they slimed their own comments section. My comment, which I posted to both Hitz' and Schwartz' blogs, was also censored only on the latter. Here it is: As others have pointed out, the main ideas behind copy-on-write and always-consistent-on-disk filesystems, such as WOFS, LSF, and Tux2, clearly predate both WAFL and ZFS. The Primäre Superblock in WOFS, for example, plays the same role as the root node in WAFL or the überblock in ZFS. Most of the other claims described in the patents are trivial applications of general-purpose data structures to filesystems. So none of the 10 patents involved in this dispute should ever have been granted. (Even to the minority of the technical community who think that software patents should be reformed rather than scrapped entirely, these particular patents are not defensible.) Both companies are trying to spin the history of this patent dispute in their favour, as you'd perhaps expect. But if you dig a bit deeper, it's worse than that -- there are inconsistencies in each story indicating that one or both must be untruthful, and my impression is that it's both. In NetApp's case, they make a cynical attempt to claim opposition to software patents despite having form in using them offensively (e.g. against BlueArc). In Sun's case, trying to extort ~$36.5M out of NetApp for the StorageTek patents (see http://www.netapp.com/go/Sun%20Lawyer%20Email.pdf) is very far from not demanding anything. Neither company comes out with any credit, and regardless of the legal outcomes, both deserve to lose from this affair in the court of public opinion. (And yes, I do mean extort. The whole patent system has degenerated into legalized extortion.) -- David Hopwood [EMAIL PROTECTED] ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] New zfs pr0n server :)))
Will Murnane wrote: On 9/6/07, Diego Righi [EMAIL PROTECTED] wrote: ...anyway I wanted to make it the most silent I could, so I suspeded all the 10 disks Warning: unfounded speculation ahead. I've heard that this can cause performance issues and undue wear on the drive. The reasoning is that since the arm assembly accelerates in one direction, and there's not much force keeping the drive from rotating, it spins in the opposite direction a little bit. This isn't a huge problem by itself, but since the place the arm was aiming for is no longer there due to the counter-rotation, it has to seek a little bit in the other direction, generating more wear and tear on the bearings, more heat from the drive, and shorter drive lifetimes. I was highly skeptical, given that drives are designed to run in both horizontal and vertical orientations (and to switch between them without reformatting, e.g. see http://www.hitachigst.com/tech/techlib.nsf/techdocs/4236D595E2C5309F862572C500813B89/$file/C10K147_IG.pdf), and that the forces on the drive arm are quite different depending on the orientation (even though the arm is counterbalanced). However, a little searching turned up http://www.sidman.com/angaccel.htm, which tends to support the reasoning above: # Accordingly, known compensation or disturbance rejection systems, while # performing satisfactorily for applications using linear or unbalanced # rotary actuators, fail to address the problems of spindle imbalance # forces, external shock or vibration and windup in systems having a # balanced rotary actuator. This failure is due to the fact that only # angular acceleration of the HDA in the direction of actuator rotation # substantially causes positioning errors in systems that utilize a balanced # rotary actuator. and suggests using an acceleration sensor specifically to compensate for angular acceleration of the HDA in the direction of actuator rotation. Suspending the drive obviously does change this acceleration. Another possible argument is that if the drive moves slightly, so does the connector to it, which seems like a bad idea. -- David Hopwood [EMAIL PROTECTED] ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Single SAN Lun presented to 4 Hosts
Rainer J.H. Brandt wrote: Ronald, thanks for your comments. I was thinking about this scenario: Host w continuously has a UFS mounted with read/write access. Host w writes to the file f/ff/fff. Host w ceases to touch anything under f. Three hours later, host r mounts the file system read-only, reads f/ff/fff, and unmounts the file system. My assumption was: a1) This scenario won't hurt w, a2) this scenario won't damage the data on the file system, a3) this scenario won't hurt r, and a4) the read operation will succeed, even if w continues with arbitrary I/O, except that it doesn't touch anything under f until after r has unmounted the file system. If the filesystem is mounted on host w, then host w is entitled to write to it at any time. If you want to reliably ensure that w does not perform any writes, then it must be unmounted on w. Note also that mounting a filesystem read-only does not guarantee that the disk will not be written, because of atime updates (this is arguably a Unix design flaw, but still has to be taken into account). So r may also write to the disk, unless the filesystem is specifically mounted with options that prevent any physical writes. Of course everything that you and Tim and Casper said is true, but I'm still inclined to try that scenario. I don't understand why you would ever want to risk this with valuable data. -- David Hopwood [EMAIL PROTECTED] ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
