Re: [zfs-discuss] "zfs send" from solaris 10/08 to "zfs receive" on solaris 10/09
Yes indeed, but my point was that it was not always so. This change was something I and others campaigned for quite some time ago. Indeed, in the early days of ZFS evangelism (when I was still at Sun) the issue came up rather often. In those days there were even more reasons not to use a zfs send stream as a backup. My point was made partly to celebrate this welcome change (as reflected in the manpage) because I was pleasantly surprised to find that the change we had requested has already been delivered. The other reason for my point was to remind people that there may be system out there for which this may not be the case, in which the old restrictions still hold until the relevant patches and/or upgrades have been applied. Indeed, the heads up link I posted seems to imply that there are still some older versions of ZFS stream format which cannot be imported - that backwards compatibility has its reasonable limits. When it come to Erik's exciting extention beyond what is already good news, perhaps the manpage needs to be stronger? It would be a shame for such useful information to be accessible only to random googlers and the august readers of this list. On 14 Nov 2009, at 17:58, Miles Nordin wrote: "ph" == Phil Harman writes: "The format of the stream is committed. You will be able to receive your streams on future versions of ZFS." What Erik said is stronger than the man page in an important way, though. He said you can dump an old stream into a filesystem on a new zpool, and when you 'zfs send' the stream back out, it'll be in the old format. Keeping this commitment means you can store s10 streams on snv_xxx backup servers as expanded filesystems, not stream-inside-a-file, and still restore them onto s10 by 'zfs send'ing from nevada. One reason this makes sense as something one might actually do because zpools are a lot more resilient to corruption than 'zfs send' streams. ___ 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/io performance on Netra X1
There is also a long-standing bug in the ALi chipset used on these servers which ZFS tickles. I don't think a work-around for this bug was ever implemented, and it's still present in Solaris 10. On Nov 13, 2009, at 11:29 AM, Richard Elling wrote: > The Netra X1 has one ATA bus for both internal drives. > No way to get high perf out of a snail. > > -- richard > > > > On Nov 13, 2009, at 8:08 AM, Bob Friesenhahn > wrote: > >> On Fri, 13 Nov 2009, Tim Cook wrote: >>> If it is using parallel SCSI, perhaps there is a problem with the SCSI bus >>> termination or a bad cable? >>> SCSI? Try PATA ;) >> >> Is that good? I don't recall ever selecting that option when purchasing a >> computer. It seemed safer to stick with SCSI than to try exotic >> technologies. >> >> Does PATA daisy-chain disks onto the same cable and controller? >> >> If this PATA and drives are becoming overwelmed, maybe it will help to tune >> zfs:zfs_vdev_max_pending down to a very small value in the kernel. >> >> Bob >> -- >> Bob Friesenhahn >> bfrie...@simple.dallas.tx.us, 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 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] Backing up ZVOLs
On Nov 14, 2009, at 1:59 PM, Brian McKerr wrote: Thanks for the help. I was curious whether the zfs send|receive was considered suitable given a few things I've read which said somethings along the lines of "don't count on being able to restore this stuff". Ideally that is what I would use with the 'incremental' option so as to only backup changed blocks on subsequent backups. Please consult the ZFS Best Practices guide. http://www.solarisinternals.com/wiki/index.php/ZFS_Best_Practices_Guide#ZFS_Backup_.2F_Restore_Recommendations -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] scrub differs in execute time?
> The ICH10 has a 32-bit/33MHz PCI bus which provides 133MB/s at half duplex. you are correct, I thought ICH10 used a 66Mhz bus, when infact its 33Mhz. The AOC card works fine in a PCI-X 64Bit/133Mhz slot good for 1,067 MB/s even if the motherboard uses a PXH chip via 8 lane PCIE. Rob ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] "zfs send" from solaris 10/08 to "zfs receive" on solaris 10/09
Miles Nordin wrote: "ph" == Phil Harman writes: >> "The format of the stream is committed. You will be able to >> receive your streams on future versions of ZFS." What Erik said is stronger than the man page in an important way, though. He said you can dump an old stream into a filesystem on a new zpool, and when you 'zfs send' the stream back out, it'll be in the old format. Keeping this commitment means you can store s10 streams on snv_xxx backup servers as expanded filesystems, not stream-inside-a-file, and still restore them onto s10 by 'zfs send'ing from nevada. One reason this makes sense as something one might actually do because zpools are a lot more resilient to corruption than 'zfs send' streams. It also came in very handy for me to archive a stream that caused a panic on Solaris 10. I was able to resend the stream to test patches. -- Ian. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Backing up ZVOLs
Thanks for the help. I was curious whether the zfs send|receive was considered suitable given a few things I've read which said somethings along the lines of "don't count on being able to restore this stuff". Ideally that is what I would use with the 'incremental' option so as to only backup changed blocks on subsequent backups. -- 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] scrub differs in execute time?
On Sat, Nov 14, 2009 at 7:00 AM, Orvar Korvar wrote: > I use Intel Q9450 + P45 Gigabyte EP45-DS3P. I put the AOC card into a PCI > slot, not PCI-x. About the HBA, I have no idea. It sounds like you're saturating the PCI port. The ICH10 has a 32-bit/33MHz PCI bus which provides 133MB/s at half duplex. This a much less than the full bandwidth from the number of drives you have on the AOC card. Getting a mobo with a PCI-X slot, getting a PCIe controller, or leaving as many drives as you can on the ICH will help performance. -B -- Brandon High : bh...@freaks.com War is Peace. Slavery is Freedom. AOL is the Internet. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] heads up on SXCE build 125 (LU + mirrored root pools)
> Seems like upgrading from b126 to b127 will have the > same problem. Yes, good point. I provided a blurb about this issue, here: http://www.solarisinternals.com/wiki/index.php/ZFS_Troubleshooting_Guide#Live_Upgrade_Problem_.28Starting_in_Nevada.2C_build_125.29 Its a good idea to review this page whenever you are considering upgrading to the next build. Cindy -- 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 send" from solaris 10/08 to "zfs receive" on solaris 10/09
> "ph" == Phil Harman writes: >> "The format of the stream is committed. You will be able to >> receive your streams on future versions of ZFS." What Erik said is stronger than the man page in an important way, though. He said you can dump an old stream into a filesystem on a new zpool, and when you 'zfs send' the stream back out, it'll be in the old format. Keeping this commitment means you can store s10 streams on snv_xxx backup servers as expanded filesystems, not stream-inside-a-file, and still restore them onto s10 by 'zfs send'ing from nevada. One reason this makes sense as something one might actually do because zpools are a lot more resilient to corruption than 'zfs send' streams. pgpe1IUEUzQ4Y.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] permanent files error, unable to access pool
I have been using opensolaris for a couple of weeks, today is my first time I reboot the system and I ran into a problem loading my external hd (meant for backup). I was expecting a more descriptive name of the file names, but given that I have no clue which ones are those, can I just tell the OS to delate those or to ignore such files? Any help would be greatly appreciated... jdrod...@_solaris:~# zpool status -v external pool: external state: ONLINE status: One or more devices has experienced an error resulting in data corruption. Applications may be affected. action: Restore the file in question if possible. Otherwise restore the entire pool from backup. see: http://www.sun.com/msg/ZFS-8000-8A scrub: none requested config: NAME STATE READ WRITE CKSUM external ONLINE 0 0 0 c10t0d0 ONLINE 0 0 0 errors: Permanent errors have been detected in the following files: :<0x0> :<0x14> jdrod...@_solaris:~# -- 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] Snapshot question
On Nov 13, 2009, at 10:36 PM, Tristan Ball wrote: I think the exception may be when doing a recursive snapshot - ZFS appears to halt IO so that it can take all the snapshots at the same instant. Snapshots cause a txg commit, similar to what you get when you run sync. The time required to commit depends on many factors, perhaps the largest of which is the latency of the disk. At least, that's what it looked like to me. I've got an Opensolaris ZFS box providing NFS to VMWare, and I was getting SCSI timeout's within the Virtual Machines that appeared to happen exactly as the snapshots were taken. SCSI timeouts?!? How short are their timeouts? By default in Solaris, SCSI timeouts are 60 seconds. Have you seen a recursive snapshot take more than 60 seconds? When I turned off the recursive snapshots, and rather had each FS snapshot individually, the problem went away. There have been performance tweeks over the past few years which can impact snapshot performance, though it is still largely gated by the disk. What release were you running? -- richard Regards, Tristan. -Original Message- From: zfs-discuss-boun...@opensolaris.org [mailto:zfs-discuss-boun...@opensolaris.org ] On Behalf Of Richard Elling Sent: Saturday, 14 November 2009 5:02 AM To: Rodrigo E. De León Plicet Cc: zfs-discuss@opensolaris.org Subject: Re: [zfs-discuss] Snapshot question On Nov 13, 2009, at 6:43 AM, Rodrigo E. De León Plicet wrote: While reading about NILFS here: http://www.linux-mag.com/cache/7345/1.html I saw this: One of the most noticeable features of NILFS is that it can "continuously and automatically save instantaneous states of the file system without interrupting service". NILFS refers to these as checkpoints. In contrast, other file systems such as ZFS, can provide snapshots but they have to suspend operation to perform the snapshot operation. NILFS doesn't have to do this. The snapshots (checkpoints) are part of the file system design itself. I don't think that's correct. Can someone clarify? It sounds to me like they confused Solaris UFS with ZFS. What they say applies to UFS, but not ZFS. -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss __ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email __ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Backing up ZVOLs
On Nov 14, 2009, at 4:38 AM, Brian McKerr wrote: Hello all, Are there any best practices / recommendations for ways of doing this ? In this case the ZVOLs would be iSCSI LUNS containing ESX VMs .I am aware of the of the need for the VMs to be quiesced for the backups to be useful. The best way is to backup closest to the context of the data -- in the VM client. But you can take snapshots of the volumes and dd or zfs send them to a backup system. Little time is required to take a snapshot, but the time required to quiesce the VM can, in my experience, be much larger than desired. Another method is to take snapshots with the VM and then save those. The same caveat applies wrt the client VM noticing the quiesce. -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] scrub differs in execute time?
On Sat, Nov 14 at 11:23, Rob Logan wrote: P45 Gigabyte EP45-DS3P. I put the AOC card into a PCI slot I'm not sure how many "half your disks" are or how your vdevs are configured, but the ICH10 has 6 sata ports at 300MB and one PCI port at 266MB (that's also shared with the IT8213 IDE chip) so in an ideal world your scrub bandwidth would be 300*6 MB with 6 disks on ICH10, in a strip 300*1 MB with 6 disks on ICH10, in a raidz 300*3+(266/3) MB with 3 disks on ICH10, and 3 on shared PCI in a strip 266/3 MB with 3 disks on ICH10, and 3 on shared PCI in a raidz 266/6 MB with 6 disks on shared PCI in a stripe 266/6 MB with 6 disks on shared PCI in a raidz we know disk don't go that fast anyway, but going from a 8h to 15h scrub is very reasonable depending on vdev config. Rob Agreed, sounds like you're saturating the PCI port. I'm pretty sure that when Thumper uses that board, they have 6 of them in PCI-X slots, which of course wouldn't have the same bandwidth limitation. --eric -- Eric D. Mudama edmud...@mail.bounceswoosh.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] [indiana-discuss] Boot failure with snv_122 and snv_123
Has this issue been solved yet? I see the same issue when trying to upgrade from 112 to 126. -- 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] scrub differs in execute time?
> P45 Gigabyte EP45-DS3P. I put the AOC card into a PCI slot I'm not sure how many "half your disks" are or how your vdevs are configured, but the ICH10 has 6 sata ports at 300MB and one PCI port at 266MB (that's also shared with the IT8213 IDE chip) so in an ideal world your scrub bandwidth would be 300*6 MB with 6 disks on ICH10, in a strip 300*1 MB with 6 disks on ICH10, in a raidz 300*3+(266/3) MB with 3 disks on ICH10, and 3 on shared PCI in a strip 266/3 MB with 3 disks on ICH10, and 3 on shared PCI in a raidz 266/6 MB with 6 disks on shared PCI in a stripe 266/6 MB with 6 disks on shared PCI in a raidz we know disk don't go that fast anyway, but going from a 8h to 15h scrub is very reasonable depending on vdev config. Rob ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] scrub differs in execute time?
I use Intel Q9450 + P45 Gigabyte EP45-DS3P. I put the AOC card into a PCI slot, not PCI-x. About the HBA, I have no idea. So I had half of the drives in the AOC card, and the other half on the mobo SATA ports. Now I have all drives to the AOC card, and suddenly a scrub takes 15h instead of 8h. Same data. This is weird. I dont get it. I dont care too much about it, but just wanted to tell you this. Thanks for your attention. -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Backing up ZVOLs
Hello all, Are there any best practices / recommendations for ways of doing this ? In this case the ZVOLs would be iSCSI LUNS containing ESX VMs .I am aware of the of the need for the VMs to be quiesced for the backups to be useful. Cheers. -- 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] moving files from one fs to another, splittin/merging
Is there a way to use only 2 or 3 digits for the second level of the var/pkg/download cache? This directory hierarchy is particularly problematic relative to moving, copying, sending, etc. This would probably speed up lookups as well. -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss