Re: [zfs-discuss] Weird performance issue with ZFS with lots of simultaneous reads
Hello Chris, Thursday, May 15, 2008, 5:42:32 AM, you wrote: CS I wrote: CS | I have a ZFS-based NFS server (Solaris 10 U4 on x86) where I am CS | seeing a weird performance degradation as the number of simultaneous CS | sequential reads increases. CS To update zfs-discuss on this: after more investigation, this seems CS to be due to file-level prefetching. Turning file-level prefetching CS off (following the directions of the ZFS Evil Tuning Guide) returns CS NFS server performance to full network bandwidth when there are lots CS of simultaneous sequential reads. Unfortunately it significantly CS reduces the performance of a single sequential read (when the server is CS otherwise idle). Have you tried to disable vdev caching and leave file level prefetching? -- Best regards, Robert Milkowskimailto:[EMAIL PROTECTED] http://milek.blogspot.com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Using O_EXCL flag on /dev/zvol nodes
Sumit Gupta wrote: The /dev/[r]dsk nodes implement the O_EXCL flag. If a node is opened using the O_EXCL, subsequent open(2) to that node fail. But I dont think the same is true for /dev/zvol/[r]dsk nodes. Is that a bug (or maybe RFE) ? Yes, that seems like a fine RFE. Or a bug, if there's a policy saying that this must work. A quick search didn't turn up any documentation. I think this would be implemented in zvol_open(). (Though it's unfortunate that every device must implement this on its own; it would be better if the infrastructure took care of it.) --matt ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] ZFS boot with gzip-9 compression
Hi, using VirtualBox I just tried to move an OpenSolaris 2008.05 boot environment (ZFS) on a gzip-9 compressed dataset, but I have the following error from grub: Error 16: Inconsistent filesystem structure Googling around I found the same error with ZFS boot and Xen in July 2007: http://mail.opensolaris.org/pipermail/xen-discuss/2007-July/000961.html http://bugs.opensolaris.org/bugdatabase/printableBug.do?bug_id=6584697 That was fixed.. Is mine another bug? BTW, using gzip-9 compression I got the whole OpenSolaris image in 815MB... The only issue is that it is not booting anymore ;-) Rgrds, Danilo. Danilo Poccia Senior Systems Engineer Sun Microsystems Italia S.p.A. Via G. Romagnosi, 4 Roma 00196 ITALY Phone +39 06 36708 022 Mobile +39 335 6983999 Fax +39 06 3221969 Email [EMAIL PROTECTED] Blog http://blogs.sun.com/danilop ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Q: ZFS Boot Install / LU support ETA?
What is the current estimated ETA on the integration of install support for ZFS boot/root support to Nevada? Also, do you have an idea when we can expect the improved ZFS write throttling to integrate? Thanks Andrew. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS boot with gzip-9 compression
Hi, using VirtualBox I just tried to move an OpenSolaris 2008.05 boot environment (ZFS) on a gzip-9 compressed dataset, but I have the following error from grub: Error 16: Inconsistent filesystem structure Googling around I found the same error with ZFS boot and Xen in July 2007: http://mail.opensolaris.org/pipermail/xen-discuss/2007-July/000961.html http://bugs.opensolaris.org/bugdatabase/printableBug.do?bug_id=6584697 That was fixed.. Is mine another bug? This is because grub does not support booting from gzip-compressed datasets now. At this moment only LZJB algorithm is supported, please see decomp_table here: http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c#65 BTW, using gzip-9 compression I got the whole OpenSolaris image in 815MB... The only issue is that it is not booting anymore ;-) Adding support for gzip may be a candidate for community project. Hth, Victor ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS boot with gzip-9 compression
Hi Victor, this seems quite easy to me but I don't know how to move around to actually implement/propose the required changes. To make grub aware of gzip (as it already is of lzjb) the steps should be: 1. create a new /onnv-gate/usr/src/grub/grub-0.95/stage2/zfs_gzip.c starting from /onnv-gate/usr/src/uts/common/fs/zfs/gzip.c and removing the gzip_compress funcion 2. add gzip_decompress at the end of /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.h 3. update the decomp_table function to link gzip and all gzip- N (with N=1...9) to the gzip_decompress function in /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c What should I do to go on with this changes? Should I start a community project? Thanks Rgrds, Danilo. Il giorno 16/mag/08, alle ore 12:53, Victor Latushkin ha scritto: Hi, using VirtualBox I just tried to move an OpenSolaris 2008.05 boot environment (ZFS) on a gzip-9 compressed dataset, but I have the following error from grub: Error 16: Inconsistent filesystem structure Googling around I found the same error with ZFS boot and Xen in July 2007: http://mail.opensolaris.org/pipermail/xen-discuss/2007-July/000961.html http://bugs.opensolaris.org/bugdatabase/printableBug.do? bug_id=6584697 That was fixed.. Is mine another bug? This is because grub does not support booting from gzip-compressed datasets now. At this moment only LZJB algorithm is supported, please see decomp_table here: http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c#65 BTW, using gzip-9 compression I got the whole OpenSolaris image in 815MB... The only issue is that it is not booting anymore ;-) Adding support for gzip may be a candidate for community project. Hth, Victor Danilo Poccia Senior Systems Engineer Sun Microsystems Italia S.p.A. Via G. Romagnosi, 4 Roma 00196 ITALY Phone +39 06 36708 022 Mobile +39 335 6983999 Fax +39 06 3221969 Email [EMAIL PROTECTED] Blog http://blogs.sun.com/danilop ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] ZFS Version Correct
I have Sun Solaris 5.10 Generic_120011-14 and the zpool version is 4. I've found references to version 5-10 on the Open Solaris site. Are these versions for Open solaris only? I've searched the SUN site for ZFS patches and found nothing (most likely operator headspace). Can I update ZFS on my Sun box and if so where are the updates? Thanks --Kenny This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
Kenny wrote: Hi! I'm new to the list and new to zfs I have the following hardware and would like opinions on implementation. Sun Enterprise T5220 FC HBA Brocade 200E 4 Gbit switch Sun 2540 FC Disk Array w/12 1TB disk drives My plan is to create a small SAN fabric with the 5220 as the initiator (additional initiators to be added later) connected to the switch and the 2540 as the target. My desire is to create 2 5disk RAID 5 sets with one hot spare each. Then using ZFS to pool the 2 sets into one 8 TB Pool with several ZFS file systems in the pool. Now I have several questions: 1) Does this plan seem ok? There doesn't seem to be anything inherently wrong with it :-) 2) Does anyone have experiance with the 2540? Kinda. I worked on adding MPxIO support to the mpt driver so we could support the SAS version of this unit - the ST2530. What sort of experience are you after? I'ver never used one of these boxes in production - only ever for benchmarking and bugfixing :-) I think Robert Milkowski might have one or two of them, however. 3) I've read that it's best practice to create the RAID set utilizing Hardware RAID utilities vice using ZFS raidz. Any wisdom on this? You've got a whacking great cache in the ST2540, so you might as well make use of it. Once you've got more questions after reading the Best Practices guide (http://www.solarisinternals.com/wiki/index.php/ZFS_Best_Practices_Guide) post a followup to this thread. You _will_ have questions. You will, I just know it! :-) cheers, James C. McPherson -- Senior Kernel Software Engineer, Solaris Sun Microsystems http://blogs.sun.com/jmcp http://www.jmcp.homeunix.com/blog ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS in S10U6 vs openSolaris 05/08
On Fri, May 16, 2008 at 09:30:27AM +0800, Robin Guo wrote: Hi, Paul At least, s10u6 will contain L2ARC cache, ZFS as root filesystem, etc.. As far as root zfs goes, are there any plans to support more than just single disks or mirrors in U6, or will that be for a later date? -brian -- Coding in C is like sending a 3 year old to do groceries. You gotta tell them exactly what you want or you'll end up with a cupboard full of pop tarts and pancake mix. -- IRC User (http://www.bash.org/?841435) ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS boot with gzip-9 compression
...just noticed there is a bug on that, but it seems there no activity even if it is in state accepted: http://bugs.opensolaris.org/view_bug.do?bug_id=6538017 Should I send an email to request-sponsor AT opensolaris DOT org to propose my fix? Rgrds, Danilo. Il giorno 16/mag/08, alle ore 14:34, Danilo Poccia ha scritto: Hi Victor, this seems quite easy to me but I don't know how to move around to actually implement/propose the required changes. To make grub aware of gzip (as it already is of lzjb) the steps should be: 1. create a new /onnv-gate/usr/src/grub/grub-0.95/stage2/zfs_gzip.c starting from /onnv-gate/usr/src/uts/common/fs/zfs/gzip.c and removing the gzip_compress funcion 2. add gzip_decompress at the end of /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.h 3. update the decomp_table function to link gzip and all gzip- N (with N=1...9) to the gzip_decompress function in /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c What should I do to go on with this changes? Should I start a community project? Thanks Rgrds, Danilo. Il giorno 16/mag/08, alle ore 12:53, Victor Latushkin ha scritto: Hi, using VirtualBox I just tried to move an OpenSolaris 2008.05 boot environment (ZFS) on a gzip-9 compressed dataset, but I have the following error from grub: Error 16: Inconsistent filesystem structure Googling around I found the same error with ZFS boot and Xen in July 2007: http://mail.opensolaris.org/pipermail/xen-discuss/2007-July/000961.html http://bugs.opensolaris.org/bugdatabase/printableBug.do?bug_id=6584697 That was fixed.. Is mine another bug? This is because grub does not support booting from gzip-compressed datasets now. At this moment only LZJB algorithm is supported, please see decomp_table here: http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c#65 BTW, using gzip-9 compression I got the whole OpenSolaris image in 815MB... The only issue is that it is not booting anymore ;-) Adding support for gzip may be a candidate for community project. Hth, Victor Danilo Poccia Senior Systems Engineer Sun Microsystems Italia S.p.A. Via G. Romagnosi, 4 Roma 00196 ITALY Phone +39 06 36708 022 Mobile +39 335 6983999 Fax +39 06 3221969 Email [EMAIL PROTECTED] Blog http://blogs.sun.com/danilop Danilo Poccia Senior Systems Engineer Sun Microsystems Italia S.p.A. Via G. Romagnosi, 4 Roma 00196 ITALY Phone +39 06 36708 022 Mobile +39 335 6983999 Fax +39 06 3221969 Email [EMAIL PROTECTED] Blog http://blogs.sun.com/danilop ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS boot with gzip-9 compression
Hello Danilo, Friday, May 16, 2008, 1:34:56 PM, you wrote: Hi Victor, this seems quite easy to me but I don't know how to "move around" to actually implement/propose the required changes. To make grub aware of gzip (as it already is of lzjb) the steps should be: 1. create a new /onnv-gate/usr/src/grub/grub-0.95/stage2/zfs_gzip.c starting from /onnv-gate/usr/src/uts/common/fs/zfs/gzip.c and removing the gzip_compress funcion 2. add gzip_decompress at the end of /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.h 3. update the decomp_table function to link "gzip" and all "gzip-N" (with N=1...9) to the gzip_decompress function in /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c What should I do to go on with this changes? Should I start a "community project"? These changes look simple enough so there is no point setting up community project imho. Just implement it, test it then ask for a sponsor and integrate it. -- Best regards, Robert mailto:[EMAIL PROTECTED] http://milek.blogspot.com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS panics solaris while switching a volume to read-only
Is there any possibility that the psarc 2007/567 can be made as a patch to Soalris 10 U5. We are planning to dispose of Veritas as quickly as possible but since all storage on production machines is on EMC Symmetrix with back-end mirroring, this panic is a showstopper for us. Or is it so intertwined that a back port of this PSARC to U5 is out of the question. Thanks Roman This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS boot with gzip-9 compression
this seems quite easy to me but I don't know how to move around to actually implement/propose the required changes. To make grub aware of gzip (as it already is of lzjb) the steps should be: 1. create a new /onnv-gate/usr/src/grub/grub-0.95/stage2/zfs_gzip.c starting from /onnv-gate/usr/src/uts/common/fs/zfs/gzip.c and removing the gzip_compress funcion Yes, but it is a little bit more complicated than that. gzip support in in-kernel ZFS leverages in-kernel zlib implementation. There is support for gzip decompression algorithm in grub (see gunzip.c), so one need to figure out how to leverage that and replace z_uncompress() with proper call. 2. add gzip_decompress at the end of /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.h 3. update the decomp_table function to link gzip and all gzip-N (with N=1...9) to the gzip_decompress function in /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c This sounds reasonable. Also one need to make sure that resulting binary does not exceed size requirements (if any), thoroughly test it to verify that it works on all HW architectures with all compression algorithms (and even mix of them). This may be not an exhaustive list of things to do. What should I do to go on with this changes? Should I start a community project? These changes look simple enough so there is no point setting up community project imho. Just implement it, test it then ask for a sponsor and integrate it. As Robert points out, this indeed may be quite simple to bother setting up community project, so it may be better to treat this just like bite-size-rfe ;-) Wbr, Victor ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
Hello James, 2) Does anyone have experiance with the 2540? JCM Kinda. I worked on adding MPxIO support to the mpt driver so JCM we could support the SAS version of this unit - the ST2530. JCM What sort of experience are you after? I'ver never used one JCM of these boxes in production - only ever for benchmarking and JCM bugfixing :-) I think Robert Milkowski might have one or two JCM of them, however. Yeah, I do have several of them (both 2530 and 2540). 2530 (SAS) - cables tend to pop-out sometimes when you are around servers... then MPxIO does not work properly if you just hot-unplug and hot-replug the sas cable... there is still 2TB LUN size limit IIRC... other than that generally it is a good value 2540 (FC) - 2TB LUN size limit IIRC, other than that it is a good value array -- Best regards, Robert Milkowskimailto:[EMAIL PROTECTED] http://milek.blogspot.com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS boot with gzip-9 compression
Hello Danilo, Friday, May 16, 2008, 2:00:42 PM, you wrote: ...just noticed there is a bug on that, but it seems there no activity even if it is in state "accepted": http://bugs.opensolaris.org/view_bug.do?bug_id=6538017 Should I send an email to request-sponsor AT opensolaris DOT org to propose my fix? Send a request and without waiting for the response just start coding :) -- Best regards, Robert mailto:[EMAIL PROTECTED] http://milek.blogspot.com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
Robert Milkowski wrote: Yeah, I do have several of them (both 2530 and 2540). 2530 (SAS) - cables tend to pop-out sometimes when you are around servers... then MPxIO does not work properly if you just hot-unplug and hot-replug the sas cable... If you plug the cable back in within 20 seconds of it coming loose that might just give MPxIO a bit of a headache. James C. McPherson -- Senior Kernel Software Engineer, Solaris Sun Microsystems http://blogs.sun.com/jmcp http://www.jmcp.homeunix.com/blog ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
On May 16, 2008, at 10:04 AM, Robert Milkowski wrote: Hello James, 2) Does anyone have experiance with the 2540? JCM Kinda. I worked on adding MPxIO support to the mpt driver so JCM we could support the SAS version of this unit - the ST2530. JCM What sort of experience are you after? I'ver never used one JCM of these boxes in production - only ever for benchmarking and JCM bugfixing :-) I think Robert Milkowski might have one or two JCM of them, however. Yeah, I do have several of them (both 2530 and 2540). we did a try and buy of the 2510,2530 and 2540. 2530 (SAS) - cables tend to pop-out sometimes when you are around servers... then MPxIO does not work properly if you just hot-unplug and hot-replug the sas cable... there is still 2TB LUN size limit IIRC... other than that generally it is a good value Yeah the sff-8088 connectors are a bit rigid and clumsy, but the performance was better than everything we tested in the 2500 series. 2540 (FC) - 2TB LUN size limit IIRC, other than that it is a good value array Echo. We like the 2540 as well, and will be buying lots of them shortly. -- Best regards, Robert Milkowskimailto:[EMAIL PROTECTED] http://milek.blogspot.com -Andy ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS in S10U6 vs openSolaris 05/08
Hi, Brian You mean stripe type with multiple-disks or raidz type? I'm afraid it's still single disk or mirrors only. If opensolaris start new project of this kind of feature, it'll be backport to s10u* eventually, but that's need some time to go, sounds no possibility in U6, I think. Brian Hechinger wrote: On Fri, May 16, 2008 at 09:30:27AM +0800, Robin Guo wrote: Hi, Paul At least, s10u6 will contain L2ARC cache, ZFS as root filesystem, etc.. As far as root zfs goes, are there any plans to support more than just single disks or mirrors in U6, or will that be for a later date? -brian ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS in S10U6 vs openSolaris 05/08
Robin Guo wrote: Hi, Brian You mean stripe type with multiple-disks or raidz type? I'm afraid it's still single disk or mirrors only. If opensolaris start new project of this kind of feature, it'll be backport to s10u* eventually, but that's need some time to go, sounds no possibility in U6, I think. Not necessarily true. Not all things in OpenSolaris get backported and not all future ZFS features are guaranteed to get backported eventually. For example I have no current plans to backport the ZFS Crypto functionality. -- Darren J Moffat ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Weird performance issue with ZFS with lots of simultaneous reads
| Have you tried to disable vdev caching and leave file level | prefetching? If you mean setting zfs_vdev_cache_bshift to 13 (per the ZFS Evil Tuning Guide) to turn off device-level prefetching then yes, I have tried turning off just that; it made no difference. If there's another tunable then I don't know about it and haven't tried it (and would be pleased to). - cks ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Hiding files with ZFS ACLs
Hi everyone, I've been experimenting with ZFS for some time and I have one question: Is it possible to hide file with ZFS ACL ? Let me explain what I would like to do: A directory (chmod 0755) contains 3 subdirs: public, private an veryprivate public has read access to everyone (0755) private has no access at all for everyone (0750), but anyone can still see it in ls veryprivate has no access at all and it doesn't even show up in ls for non authorized user (let say non-root for simplification). I tried removing read_acl for everyone on veryprivate, but ls shows a error: ls: can't read ACL on ./veryprivate: Permission denied Is it possible to do this with acls ? Thank in advance, Vincent This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS boot with gzip-9 compression
I have been using zfs boot with lzjb compression on since build 75, from time to time I had similar problem, not sure why. As best practise, I do snapshot the root filesystem frequently, so that I can rollback to the last working snapshot. Rgds, Andre W. Victor Latushkin wrote: this seems quite easy to me but I don't know how to move around to actually implement/propose the required changes. To make grub aware of gzip (as it already is of lzjb) the steps should be: 1. create a new /onnv-gate/usr/src/grub/grub-0.95/stage2/zfs_gzip.c starting from /onnv-gate/usr/src/uts/common/fs/zfs/gzip.c and removing the gzip_compress funcion Yes, but it is a little bit more complicated than that. gzip support in in-kernel ZFS leverages in-kernel zlib implementation. There is support for gzip decompression algorithm in grub (see gunzip.c), so one need to figure out how to leverage that and replace z_uncompress() with proper call. 2. add gzip_decompress at the end of /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.h 3. update the decomp_table function to link gzip and all gzip-N (with N=1...9) to the gzip_decompress function in /onnv-gate/usr/src/grub/grub-0.95/stage2/fsys_zfs.c This sounds reasonable. Also one need to make sure that resulting binary does not exceed size requirements (if any), thoroughly test it to verify that it works on all HW architectures with all compression algorithms (and even mix of them). This may be not an exhaustive list of things to do. What should I do to go on with this changes? Should I start a community project? These changes look simple enough so there is no point setting up community project imho. Just implement it, test it then ask for a sponsor and integrate it. As Robert points out, this indeed may be quite simple to bother setting up community project, so it may be better to treat this just like bite-size-rfe ;-) Wbr, Victor ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
On Fri, 16 May 2008, Kenny wrote: Sun 2540 FC Disk Array w/12 1TB disk drives It is interesting that the 2540 is available with large disks now. My desire is to create 2 5disk RAID 5 sets with one hot spare each. Then using ZFS to pool the 2 sets into one 8 TB Pool with several ZFS file systems in the pool. Now I have several questions: 1) Does this plan seem ok? Another option is to export each entire drive as a LUN and put 10 active drives into one zfs pool as two raidzs, or one raidz2. The other two drives can be retained as spares for the pool. If replacement drives are readily sourced on demand, you could use all 12 drives as two raidz2s. This approach does not allow other systems to use the 2540 since one host owns the pool. However, you could move the ZFS pool to another system if need be. 2) Does anyone have experiance with the 2540? Yes. Please see my white paper at http://www.simplesystems.org/users/bfriesen/zfs-discuss/2540-zfs-performance.pdf which discusses my experience with ZFS and the 2540. The paper was written back in February and I have yet to experience a hickup with the 2540 or ZFS. Not even one bad block. 3) I've read that it's best practice to create the RAID set utilizing Hardware RAID utilities vice using ZFS raidz. Any wisdom on this? This is really a philosophical or requirements issue. The 2540 allows you create pools and then export only part of the pool as a LUN to be used by an initator. This allows you create LUNs on disks which are shared by multiple hosts (initiators), each of which has its own ZFS pool (or traditional filesystem). If you really need to divide up storage at this level, then the 2540 offers flexibility that you won't get from ZFS. A drawback to sharing sliced pools in this way is that if there is a problem with the underlying disks, then multiple hosts may be impacted during recovery. The 2540 CAM provides a 4-disk RAID5 config which claims to be tuned for ZFS. Someone on the list created three 4-disk RAID5 LUNs this way and put them all in one ZFS pool, obtaining very good performance. If one of those LUNs was to irreparably fail, his entire pool would be toast. ZFS experts will tell you that you should not be trusting the 2540 or its firmware to catch all errors and so there should always be redundancy (e.g. mirroring) at the ZFS level. By exporting each 2540 disk as a LUN, then any of the redundancy schemes supported by ZFS (mirror, raidz, raidz2) can be used from the initiator, essentially ignoring the ones built into the 2540. While the 2540's CAM interface is nice, you will find that it is far slower than ZFS is at incorporating your disks (25 tedious minutes in the CAM admin tool vs less than a second for ZFS). Bob == Bob Friesenhahn [EMAIL PROTECTED], http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer,http://www.GraphicsMagick.org/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
On Fri, 16 May 2008, James C. McPherson wrote: 3) I've read that it's best practice to create the RAID set utilizing Hardware RAID utilities vice using ZFS raidz. Any wisdom on this? You've got a whacking great cache in the ST2540, so you might as well make use of it. Exporting each disk as a LUN for use by ZFS does not cause the 2540 to disable its cache. In fact, it is clear that this cache is quite valuable to ZFS write performance when NFS is involved. I am able to obtain 90MB/second NFS write performance from a single NFS client and using the 2540. Due to the inherent design of ZFS, it is not necessary for RAID writes to be synchronized as they must be for traditional mirroring or RAID5. If there is a power loss or crash, ZFS will discover where it left off, and bring all redundant copies to a coherent state. The 2540's cache will help protect against losing data if there is a power fail. Bob == Bob Friesenhahn [EMAIL PROTECTED], http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer,http://www.GraphicsMagick.org/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] ZFS sharing question.
Hello. Anyone out there remember the -d option for share? How do you set the share description using the zfs set commands, or is it even possible? Thanks! -B This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] cp -p gives errors on Linux w/ NFS-mounted ZFS
Hello all, I'm having the same problem here, any news? I need to use ACL's on the GNU/Linux clients. I'm using nfsv3, and on the GNU/Linux servers that feature was working, i think we need a solution for solaris/opensolaris. Now, with the dmm project, how we can start a migration process, if we can not provide the same services on the target machine? Thanks a lot for your time! Leal. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
I run 3510FC and 2540 units in pairs. I build 2 5-disk RAID5 LUNs in each array, with 2 disks as global spares. Each array has dual controllers and I'm doing multipath. Then from the server I have access to 2 LUNs from 2 arrays, and I build a ZFS RAID-10 set from these 4 LUNs being sure each mirror pair is constructed with LUNs from both arrays. Thus I can survive a complete failure of one array and multiple other failures and keep on trucking. Performance is quite good since using this in /etc/system: set zfs:zfs_nocacheflush = 1 And since recent ZFS patches for 10u4 which fixed FSYNC performance issues my arrays and servers are hardly breaking a sweat. I very much like that the arrays can handle lower-level problems for me like sparing and ZFS ensures correctness on top of that. This is for Cyrus mail-stores so availability and correctness are paramount in case you are wondering if all this belt suspenders paranoia is worthwhile. If/when ZFS acquires a method to ensure that spare#1 in chassis#1 only gets used to replace failed disks in chassis#1 then I'll reconsider my position. Currently though there is no mechanism to ensure this so I could easily see a spare being pulled from the other chassis and leaving me with an undesirable dependency if I were doing ZFS with JBOD. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
On Fri, 16 May 2008, Vincent Fox wrote: If/when ZFS acquires a method to ensure that spare#1 in chassis#1 only gets used to replace failed disks in chassis#1 then I'll reconsider my position. Currently though there is no mechanism to ensure this so I could easily see a spare being pulled from the other chassis and leaving me with an undesirable dependency if I were doing ZFS with JBOD. Good point! However, I think that the spare is only used until the original is re-constructed, so its usage should not be very long. Bob == Bob Friesenhahn [EMAIL PROTECTED], http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer,http://www.GraphicsMagick.org/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
Actually, I only meant that zfs boot was integrated into build 90. I don't know about the improved write throttling. I will check into why there was no mention of this on the heads up page. Lori Andrew Pattison wrote: Were both of these items (ZFS boot install support and improved write throttling) integrated into build 90? I don't see any mention of this on the Nevada head up page. Thanks Andrew. On Fri, May 16, 2008 at 5:21 PM, Lori Alt [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: It has been integrated into Nevada build 90. Lori andrew wrote: What is the current estimated ETA on the integration of install support for ZFS boot/root support to Nevada? Also, do you have an idea when we can expect the improved ZFS write throttling to integrate? Thanks Andrew. This message posted from opensolaris.org http://opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org mailto:zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Andrew Pattison andrum04 at gmail dot com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
Clarifying further: the install support for zfs root file systems went into build 90, but because the current install code is closed source, the effect of that integration will not be seen until the build 90 SXCE is released. At that time, installs will show a screen that give the user an opportunity to choose between installing a ufs or a zfs root on the system (ufs is still the default). Lori Lori Alt wrote: Actually, I only meant that zfs boot was integrated into build 90. I don't know about the improved write throttling. I will check into why there was no mention of this on the heads up page. Lori Andrew Pattison wrote: Were both of these items (ZFS boot install support and improved write throttling) integrated into build 90? I don't see any mention of this on the Nevada head up page. Thanks Andrew. On Fri, May 16, 2008 at 5:21 PM, Lori Alt [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: It has been integrated into Nevada build 90. Lori andrew wrote: What is the current estimated ETA on the integration of install support for ZFS boot/root support to Nevada? Also, do you have an idea when we can expect the improved ZFS write throttling to integrate? Thanks Andrew. This message posted from opensolaris.org http://opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org mailto:zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Andrew Pattison andrum04 at gmail dot com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
Install of a zfs root can only be done with the tty-based installer or with Jumpstart. I will make sure that instructions for both are made available by the time that SXDE build 90 is released. For an attractive, easy-to-use installer that is designed from the outset to install systems with zfs root pools, we'll all have to wait for Caiman: http://www.opensolaris.org/os/project/caiman/ Lori Andrew Pattison wrote: Just to clarify, if I run the old Java-based installer for Solaris Express on build 90, it will allow me to install Solaris to a ZFS pool? Thanks Andrew. On Fri, May 16, 2008 at 9:03 PM, Lori Alt [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Actually, I only meant that zfs boot was integrated into build 90. I don't know about the improved write throttling. I will check into why there was no mention of this on the heads up page. Lori Andrew Pattison wrote: Were both of these items (ZFS boot install support and improved write throttling) integrated into build 90? I don't see any mention of this on the Nevada head up page. Thanks Andrew. On Fri, May 16, 2008 at 5:21 PM, Lori Alt [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: It has been integrated into Nevada build 90. Lori andrew wrote: What is the current estimated ETA on the integration of install support for ZFS boot/root support to Nevada? Also, do you have an idea when we can expect the improved ZFS write throttling to integrate? Thanks Andrew. This message posted from opensolaris.org http://opensolaris.org http://opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org mailto:zfs-discuss@opensolaris.org mailto:zfs-discuss@opensolaris.org mailto:zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Andrew Pattison andrum04 at gmail dot com -- Andrew Pattison andrum04 at gmail dot com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
On Fri, May 16, 2008 at 02:32:34PM -0600, Lori Alt wrote: Install of a zfs root can only be done with the tty-based installer or with Jumpstart. I will make sure that instructions for both are made available by the time that SXDE build 90 is released. Will the tty or jumpstart based installed be able to do an Upgrade install over an older ZFS Root system? For an attractive, easy-to-use installer that is designed from the outset to install systems with zfs root pools, we'll all have to wait for Caiman: We don't need no stinkin' attractive, easy-to-use installers. ;) http://www.opensolaris.org/os/project/caiman/ Is the tty version going to stick around? Or will Caiman have a tty version? -brian -- Coding in C is like sending a 3 year old to do groceries. You gotta tell them exactly what you want or you'll end up with a cupboard full of pop tarts and pancake mix. -- IRC User (http://www.bash.org/?841435) ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
So it's pushed back to build 90 now? There was an announcement the other day that build 88 was being skipped and build 89 would be the official release with ZFS boot. Not a big deal but someone should do an announcement about the change. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
By my calculations that makes the possible release date for ZFS boot installer support around the 9th June 2008. Mark that date in your diary! Cheers Andrew. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS in S10U6 vs openSolaris 05/08
The issues with CIFS is not just complexity; it's the total amount of incompatible change in the kernel that we had to make in order to make the CIFS protocol a first class citizen in Solaris. This includes changes in the VFS layer which would break all S10 file systems. So in a very real sense CIFS simply cannot be backported to S10. -- Fred On May 16, 2008, at 3:06 PM, Paul B. Henson wrote: On Thu, 15 May 2008, Robin Guo wrote: The most feature and bugfix so far towards Navada 87 (or 88? ) will backport into s10u6. It's about the same (I mean from outside viewer, not inside) with openSolaris 05/08, but certainly, some other features as CIFS has no plan to backport to s10u6 yet, so ZFS will has fully ready but no effect on these kind of area. That depend on how they co- operate. Yah, I've heard that the CIFS stuff was way too many changes to backport, guess that is going to have to wait until Solaris 11. So, from a feature perspective it looks like S10U6 is going to be in pretty good shape ZFS-wise. If only someone could speak to (perhaps under the cloak of anonymity ;) ) the timing side :). Given U5 barely came out, I wouldn't expect U6 anytime soon :(. Thanks... -- Paul B. Henson | (909) 979-6361 | http://www.csupomona.edu/ ~henson/ Operating Systems and Network Analyst | [EMAIL PROTECTED] California State Polytechnic University | Pomona CA 91768 ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Fred Zlotnick Senior Director, Solaris NAS Sun Microsystems, Inc. [EMAIL PROTECTED] x81142/+1 650 352 9298 ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
Brian Hechinger wrote: On Fri, May 16, 2008 at 02:32:34PM -0600, Lori Alt wrote: Install of a zfs root can only be done with the tty-based installer or with Jumpstart. I will make sure that instructions for both are made available by the time that SXDE build 90 is released. Will the tty or jumpstart based installed be able to do an Upgrade install over an older ZFS Root system? Sadly, no. You can bfu to the Solaris build 88 bits, but it's probably best to reinstall with the new installer when it becomes available. For an attractive, easy-to-use installer that is designed from the outset to install systems with zfs root pools, we'll all have to wait for Caiman: We don't need no stinkin' attractive, easy-to-use installers. ;) http://www.opensolaris.org/os/project/caiman/ Is the tty version going to stick around? Or will Caiman have a tty version? The tty installer you will see in Build 90 will be going away at some point. It's clunky and written for the days when a 400 MB disk was considered really big. It's very ufs-centric and didn't adapt well to the idea of storage pools. I can't say for sure what Caiman will provide but I notice this among its goals, as specified on its OpenSolaris community web page: * Updated and simplified graphical and text user interfaces which carry Sun's current branding - Lori ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS in S10U6 vs openSolaris 05/08
On Fri, May 16, 2008 at 03:12:02PM -0700, Zlotnick Fred wrote: The issues with CIFS is not just complexity; it's the total amount of incompatible change in the kernel that we had to make in order to make the CIFS protocol a first class citizen in Solaris. This includes changes in the VFS layer which would break all S10 file systems. So in a very real sense CIFS simply cannot be backported to S10. However, the same arguments were made explaining the difficulty backporting ZFS and GRUB boot to Solaris 10. Adam -- Adam Leventhal, Fishworkshttp://blogs.sun.com/ahl ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
On Fri, May 16, 2008 at 02:19:29PM -0600, Lori Alt wrote: Clarifying further: the install support for zfs root file systems went into build 90, but because the current install code is closed source, the effect of that integration will not be seen until the build 90 SXCE is released. At that time, installs will show a screen that give the user an opportunity to choose between installing a ufs or a zfs root on the system (ufs is still the default). I'm hoping that the build 90 WOS images, when they are posted, will be enough. More importantly, I'm hoping that I can get OpenSolaris 2008.05 and Solaris Nevada build 90 to co-exist on the same root pool. Nico -- ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
On Fri, May 16, 2008 at 01:59:56PM -0700, Vincent Fox wrote: So it's pushed back to build 90 now? Evidently, but build 90 is closed, and the bits are in. The WOS images for build 90 are not out yet, but that's a matter of time; the bits are in. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Q: ZFS Boot Install / LU support ETA?
The write throttling improvement is in build 87. -- Prabahar. Lori Alt wrote: Actually, I only meant that zfs boot was integrated into build 90. I don't know about the improved write throttling. I will check into why there was no mention of this on the heads up page. Lori Andrew Pattison wrote: Were both of these items (ZFS boot install support and improved write throttling) integrated into build 90? I don't see any mention of this on the Nevada head up page. Thanks Andrew. On Fri, May 16, 2008 at 5:21 PM, Lori Alt [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: It has been integrated into Nevada build 90. Lori andrew wrote: What is the current estimated ETA on the integration of install support for ZFS boot/root support to Nevada? Also, do you have an idea when we can expect the improved ZFS write throttling to integrate? Thanks Andrew. This message posted from opensolaris.org http://opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org mailto:zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Andrew Pattison andrum04 at gmail dot com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS panics solaris while switching a volume to read-only
The fix is already in Solaris 10 U6. A patch for S10U5 will only be available when S10U6 is released. -- Prabahar. Veltror wrote: Is there any possibility that the psarc 2007/567 can be made as a patch to Soalris 10 U5. We are planning to dispose of Veritas as quickly as possible but since all storage on production machines is on EMC Symmetrix with back-end mirroring, this panic is a showstopper for us. Or is it so intertwined that a back port of this PSARC to U5 is out of the question. Thanks Roman This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
Bob Friesenhahn wrote: On Fri, 16 May 2008, James C. McPherson wrote: 3) I've read that it's best practice to create the RAID set utilizing Hardware RAID utilities vice using ZFS raidz. Any wisdom on this? You've got a whacking great cache in the ST2540, so you might as well make use of it. Exporting each disk as a LUN for use by ZFS does not cause the 2540 to disable its cache. In fact, it is clear that this cache is quite valuable to ZFS write performance when NFS is involved. I am able to obtain 90MB/second NFS write performance from a single NFS client and using the 2540. Due to the inherent design of ZFS, it is not necessary for RAID writes to be synchronized as they must be for traditional mirroring or RAID5. If there is a power loss or crash, ZFS will discover where it left off, and bring all redundant copies to a coherent state. The 2540's cache will help protect against losing data if there is a power fail. Hi Bob, You've made an assumption about what I wrote. That assumption is incorrect. Kenny, in addition, did not say that he was or was not going to do what you suggested, and I suggested to him that he go and look into the ZFS Best Practices wiki to get some ideas. I'm very, very well aware of the design and behaviour of ZFS, I have been using it since the build it was first integrated. I am also quite well aware of the design and behaviour of the raid engine in the ST2530. Please re-read my email to Kenny, and don't put in words that I didn't write. James C. McPherson -- Senior Kernel Software Engineer, Solaris Sun Microsystems http://blogs.sun.com/jmcp http://www.jmcp.homeunix.com/blog ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
On Sat, 17 May 2008, James C. McPherson wrote: Bob Friesenhahn wrote: On Fri, 16 May 2008, James C. McPherson wrote: 3) I've read that it's best practice to create the RAID set utilizing Hardware RAID utilities vice using ZFS raidz. Any wisdom on this? You've got a whacking great cache in the ST2540, so you might as well make use of it. Hi Bob, You've made an assumption about what I wrote. That assumption is incorrect. Kenny, in addition, did not say that he was or My assumption based on your You've got a whacking great cache in the ST2540, so you might as well make use of it. was that it was intended to imply that if the Hardware RAID utilities were not used that the 2540's NV write-cache would not be available/useful. I am also quite well aware of the design and behaviour of the raid engine in the ST2530. Since there seems to be no specification of the internal architecture of the 2530 and 2540 (quite odd for a Sun product!), perhaps you can create a whitepaper which describes this architecture so that Sun customers can better understand how to use the product. I have nothing but praise for the two Sun engineers who helped me understand and optimize for the 2540 back in February. Most Sun engineers on this list are very helpful and we are very thankful for their kind assistance. Bob == Bob Friesenhahn [EMAIL PROTECTED], http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer,http://www.GraphicsMagick.org/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and Sun Disk arrays - Opinions?
Bob Friesenhahn wrote: On Sat, 17 May 2008, James C. McPherson wrote: Bob Friesenhahn wrote: On Fri, 16 May 2008, James C. McPherson wrote: 3) I've read that it's best practice to create the RAID set utilizing Hardware RAID utilities vice using ZFS raidz. Any wisdom on this? You've got a whacking great cache in the ST2540, so you might as well make use of it. Hi Bob, You've made an assumption about what I wrote. That assumption is incorrect. Kenny, in addition, did not say that he was or My assumption based on your You've got a whacking great cache in the ST2540, so you might as well make use of it. was that it was intended to imply that if the Hardware RAID utilities were not used that the 2540's NV write-cache would not be available/useful. Indeed. And there is absolutely no justification for that assumption. I had hoped that the following sentence suggesting a perusal of the Best Practices would have made it clear that it is indeed possible (I would say, recommended) to maximise the usage of a cache and ZFS' specific design features. I am also quite well aware of the design and behaviour of the raid engine in the ST2530. Since there seems to be no specification of the internal architecture of the 2530 and 2540 (quite odd for a Sun product!), perhaps you can create a whitepaper which describes this architecture so that Sun customers can better understand how to use the product. I'm not the person to do that, but I will forward your suggestion on to somebody who is better placed to do so. James C. McPherson -- Senior Kernel Software Engineer, Solaris Sun Microsystems http://blogs.sun.com/jmcp http://www.jmcp.homeunix.com/blog ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] openSolaris ZFS root, swap, dump
Historically I've used hardware raid1 for the boot disks on my servers. With the availability of ZFS root, I want to explore making the two underlying drives directly available to the operating system and create a ZFS mirror to avail of error detection and self-healing. The current openSolaris installer, when given an entire disk, partitions it into two slices: s0 for the ZFS pool, and s1 for swap. Once the system is installed, I could add the second disk as a mirror, giving redundancy for the pool, but that would leave swap as a single point of failure. I investigated using a ZFS zvol for swap, which appears to be a viable option, but based on what I read you could not use a ZFS zvol for a dump device. The dumpadm man page under openSolaris says: For systems with a ZFS root file system, dedicated ZFS volumes are used for swap and dump areas. For further information about setting up a dump area with ZFS, see the ZFS Administration Guide. However, the actual documentation ahttp://docs.sun.com/app/docs/doc/819-5461/gaypf?l=en' says Using a ZFS volume as a dump device is not supported. That is Solaris 10 documentation, though. Can openSolaris actually use a ZFS zvol for dump? A somewhat related question, I was trying to determine the best way to convert the default install into the disk layout I want. For ZFS root, is it required to have a partition and slices? Or can I just give it the whole disk and have it write an EFI label on it? I know you cannot remove disks currently from a pool, but you can swap out devices, correct? So can I swap in the second disk (either the whole disk, or the overlap partition of a legacy labeled disk) to replace the initial installation disk, and then attach the original disk again as the second half of the mirror (again either the entire disk, or the overlap slice)? -- Paul B. Henson | (909) 979-6361 | http://www.csupomona.edu/~henson/ Operating Systems and Network Analyst | [EMAIL PROTECTED] California State Polytechnic University | Pomona CA 91768 ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] openSolaris ZFS root, swap, dump
Paul B. Henson wrote: Historically I've used hardware raid1 for the boot disks on my servers. With the availability of ZFS root, I want to explore making the two underlying drives directly available to the operating system and create a ZFS mirror to avail of error detection and self-healing. The current openSolaris installer, when given an entire disk, partitions it into two slices: s0 for the ZFS pool, and s1 for swap. This is because OpenSolaris 2008.05 is based on NV b86 which does not have the fix for 5008936 ZFS and/or zvol should support dumps 5070124 dumpadm -d /dev/... does not enforce block device requirement for savecore 6633197 zvol should not permit newfs or createpool while it's in use by swap or dump which were integrated into NV b88 as part of ZFS boot support, http://opensolaris.org/os/community/on/flag-days/pages/2008041103 As they say, timing is everything... -- richard Once the system is installed, I could add the second disk as a mirror, giving redundancy for the pool, but that would leave swap as a single point of failure. I investigated using a ZFS zvol for swap, which appears to be a viable option, but based on what I read you could not use a ZFS zvol for a dump device. The dumpadm man page under openSolaris says: For systems with a ZFS root file system, dedicated ZFS volumes are used for swap and dump areas. For further information about setting up a dump area with ZFS, see the ZFS Administration Guide. However, the actual documentation ahttp://docs.sun.com/app/docs/doc/819-5461/gaypf?l=en' says Using a ZFS volume as a dump device is not supported. That is Solaris 10 documentation, though. Can openSolaris actually use a ZFS zvol for dump? A somewhat related question, I was trying to determine the best way to convert the default install into the disk layout I want. For ZFS root, is it required to have a partition and slices? Or can I just give it the whole disk and have it write an EFI label on it? I know you cannot remove disks currently from a pool, but you can swap out devices, correct? So can I swap in the second disk (either the whole disk, or the overlap partition of a legacy labeled disk) to replace the initial installation disk, and then attach the original disk again as the second half of the mirror (again either the entire disk, or the overlap slice)? ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss