Re: [zfs-discuss] MPT questions
Frank Middleton wrote: In an attempt to recycle some old PATA disks, we bought some really cheap PATA/SATA adapters, some of which actually work to the point where it is possible to boot from a ZFS installation (e.g., c1t2d0s0). Not all PATA disks work, just Seagates, it would seem, but not Maxstors. I wonder why? probe-scsi-all sees Seagate but not Maxstor disks plugged into the same adapter. Such disks have proven invaluable as a substitute for rescue CDs until such CDs become possible. The odd thing is that booting from another disk, ZFS can't see the adapted disk even though it is bootable. Could the reason be that there's no /dev/rdsk/c1t2d0, but there are c1t0d0, etc.? Format sees the disk but zpool import doesn't (this is on SPARC sun4u). This isn't at all important, just curious as to why this might be and why zpool import can't see it at all, but zpool create can. Gotta say how happy we are with the MPT driver and the LSI SAS controller - fast and reliable - petabytes of i/o and not a single zfs checksum error! This has little to do with ZFS, but should it be possible to see a PATA CD or DVD connected to an MPT (LSI) SAS controller via one of these adapters? Though I'd ask before forking out for a SATA DVD drive - just hate to put perfectly good drives out for recycling. It might work. It certainly wouldn't hurt to try. 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] MPT questions
In an attempt to recycle some old PATA disks, we bought some really cheap PATA/SATA adapters, some of which actually work to the point where it is possible to boot from a ZFS installation (e.g., c1t2d0s0). Not all PATA disks work, just Seagates, it would seem, but not Maxstors. I wonder why? probe-scsi-all sees Seagate but not Maxstor disks plugged into the same adapter. Such disks have proven invaluable as a substitute for rescue CDs until such CDs become possible. The odd thing is that booting from another disk, ZFS can't see the adapted disk even though it is bootable. Could the reason be that there's no /dev/rdsk/c1t2d0, but there are c1t0d0, etc.? Format sees the disk but zpool import doesn't (this is on SPARC sun4u). This isn't at all important, just curious as to why this might be and why zpool import can't see it at all, but zpool create can. Gotta say how happy we are with the MPT driver and the LSI SAS controller - fast and reliable - petabytes of i/o and not a single zfs checksum error! This has little to do with ZFS, but should it be possible to see a PATA CD or DVD connected to an MPT (LSI) SAS controller via one of these adapters? Though I'd ask before forking out for a SATA DVD drive - just hate to put perfectly good drives out for recycling. Maybe someone can recommend a writable BlueRay SAS drive that is known to work with the MPT driver instead... Thanks -- Frank ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Hot Spares spin down?
Hi there, On Oct 8, 2009, at 9:46 PM, bjbm wrote: Sorry if this is a noob question but I can't seem to find this info anywhere. Are hot spares generally spunned down until they are needed? No, but have a look at power.conf(4) and the device-thresholds keyword to spin down disks. Here is a bigadmin article also: http://www.sun.com/bigadmin/features/articles/disk_power_saving.jsp Regards Henrik http://sparcv9.blogspot.com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] ZFS saved my data success story
To recap for those who don't recall my plaintive cries for help, I lost a pool due to the following sequence of events: - One drive in my raidz array becomes flaky, has frequent "stuck" I/Os due to drive error recovery, trashing performance - I take flaky drive offline (zpool offline...) - I bring down the server and swap out the disk - When the server comes back up, the pool won't import, because a _different_ drive has decided to not spin up - I swap the "flaky" disk back in - Pool won't import because the "flaky" disk is several TxGs behind (and, if I boot into a boot CD, because it's "offline") Victor Latushkin from Sun's Moscow office once again comes to the rescue, rolling back the uberblocks to a (mostly) sane TxG ID, and providing new zfs bits with a read-only import option (to make sure I didn't trash things worse than they already were...) I have finally completed recovery of my 3+ TB of data. I lost one email account, about 10 non-spam non-trash emails from my email account, and one ISO image. I'm fairly certain that rolling back another few TxGs would have gotten the lost email account directory back, but as the user had a complete offline copy on his desktop, it wasn't worth it. I suspect the emails that were not recoverable from my account were actually ones I had recently deleted / filed as spam, but I can't be certain (I know the file names, but 1013456. isn't very descriptive...) If I had been using any other file system, I would have lost everything. I had partial backups of my most critical data, but the loss would have been extremely painful. Due to the design of ZFS and the help from Victor (hopefully soon to no longer be needed with delivery of the zpool rollback tool(s)), not only did I recover the vast majority of my data, I know what I didn't recover, and I know that what I did recover is not corrupt. Thanks, ZFS! I am _so_ glad I became an early adopter many years ago when I built this home server. For the record, my new, larger pool is raidz2, and I am plugging the holes in my offsite backup syncs to cover more of my data (not all of it, as 8+ TB of data is a lot... I will probably purchase some big, slow disks next year to allow me to keep a full offline backup locally). -- Carson ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Hot Spares spin down?
Sorry if this is a noob question but I can't seem to find this info anywhere. Are hot spares generally spunned down until they are needed? -- 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] scrubing/resilvering - controller problem
You might try setting zfs_scrub_limit to 1 or 2 and attach a customer service record to : 6494473 ZFS needs a way to slow down resilvering -r Le 7 oct. 09 à 06:14, John a écrit : Hi, We are running b118, with a LSI 3801 controller which is connected to 44 drives (yes it's a lot behind a single controller). We also use a pair of ssd connected to another controller for read cache. Everything works fine and we achieve acceptable performance for our needs. However, during scrubbing or resilvering operations, it seems ZFS generates so much traffic that it overwhelmes the controller. The controller then logs the following errors: Oct 6 07:30:04 nas101 scsi: [ID 107833 kern.warning] WARNING: / p...@0,0/pci8086,6...@4/pci1000,3...@0/s...@16,0 (sd19): Oct 6 07:30:04 nas101incomplete read- retrying Is there anything that can be done to slow down zfs operations such as resilvering/scrubbing? We tried tuning zfs:zfs_vdev_max_pending but it did not really help. This is a bit frustrating because this configuration works well to serve data. It's just too aggressive when the kernel accesses drives for some operations. Iostat looks liks this: r/sw/s kr/s kw/s wait actv wsvc_t asvc_t %w %b s/w h/w trn tot device 9681.5 37.3 116403.3 35.1 0.0 30.20.03.1 0 1000 0 21 8 29 c9 420.31.5 5058.21.4 0.0 1.30.03.1 0 44 0 0 0 0 c9t8d0 Any help would be appreciated. thanks, JJ -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss smime.p7s Description: S/MIME cryptographic signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] zfs send/receive performance concern
I am running zfs send/receive on a ~1.2Tb zfs spread across 10x200Gb LUNs. Has copied only 650Gb in ~42Hrs. Source pool and destination pool are from the same storage sub system. Last time when ran, took ~20Hrs. Something is terribly wrong here. What do i need to look to figure out the reason? ran zpool iostat and iostat on the given pool for some clue. but still in a state of confusion now. RR. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] convert raidz from osx
yes, that was what I was doing. I wanted to give the raidz whole disks because grub didn't want to install. ( I forgot which command I used . bootadm ? ) I have another slice free on a shared disk with osx and win7 but I am having problems with grub. I will try that again and document it so I can ask a proper question about it. As the installer has problems and I couldn't get the fdisk workaround to work with the osol-1002-118-x86.iso Dirk On 08 Oct 2009, at 16:03, Cindy Swearingen wrote: Dirk, I'm not sure I'm following you exactly but this is what I think you are trying to do: You have a RAIDZ pool that is built with slices and you are trying to convert the slice configuration to whole disks. This isn't possible because you are trying replace the same disk. This is what happens: # zpool create test raidz c0t4d0s0 c0t5d0s0 c0t6d0s0 # zpool replace test c0t6d0s0 c0t6d0 invalid vdev specification use '-f' to override the following errors: /dev/dsk/c0t6d0s0 is part of active ZFS pool test. Please see zpool (1M). You could replace the disk slice with a different disk like this: # zpool replace test c0t6d0s0 c0t7d0 If you don't have any additional disks then I think you will have to backup the data and recreate the pool. Maybe someone else has a better idea. Also, you refer to rpool, which is the default name of the ZFS root pool in the Opensolaris release. This pool cannot be RAIDZ pool nor can it contain whole disks. It must be created with disk slices. Cindy On 10/08/09 02:15, dirk schelfhout wrote: I am converting a 4 disk raidz from osx to opensolaris. And I want to keep the data intact. I want zfs to get access to the full disk instead of a slice. I believe like c8d0 instead off c8d0s1. I wanted to do this 1 disk at a time and let it resilver. what is the proper way to do this. I tried, I believe from memory : zpool replace -f rpool c8d1s1 c8d1 but it didn't let me do that. then I tried to put the disk offline first , but same result. Thanks, Dirk ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] convert raidz from osx
Dirk, I'm not sure I'm following you exactly but this is what I think you are trying to do: You have a RAIDZ pool that is built with slices and you are trying to convert the slice configuration to whole disks. This isn't possible because you are trying replace the same disk. This is what happens: # zpool create test raidz c0t4d0s0 c0t5d0s0 c0t6d0s0 # zpool replace test c0t6d0s0 c0t6d0 invalid vdev specification use '-f' to override the following errors: /dev/dsk/c0t6d0s0 is part of active ZFS pool test. Please see zpool(1M). You could replace the disk slice with a different disk like this: # zpool replace test c0t6d0s0 c0t7d0 If you don't have any additional disks then I think you will have to backup the data and recreate the pool. Maybe someone else has a better idea. Also, you refer to rpool, which is the default name of the ZFS root pool in the Opensolaris release. This pool cannot be RAIDZ pool nor can it contain whole disks. It must be created with disk slices. Cindy On 10/08/09 02:15, dirk schelfhout wrote: I am converting a 4 disk raidz from osx to opensolaris. And I want to keep the data intact. I want zfs to get access to the full disk instead of a slice. I believe like c8d0 instead off c8d0s1. I wanted to do this 1 disk at a time and let it resilver. what is the proper way to do this. I tried, I believe from memory : zpool replace -f rpool c8d1s1 c8d1 but it didn't let me do that. then I tried to put the disk offline first , but same result. Thanks, Dirk ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] strange pool disks usage pattern
By the way, there are more than fifty bugs logged for marevell88sx, many of them about problems with DMA handling and/or driver behaviour under stress. Can it be that I'm stumbling upon something along these lines? http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6826483 Maurilio. -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] convert raidz from osx
I am converting a 4 disk raidz from osx to opensolaris. And I want to keep the data intact. I want zfs to get access to the full disk instead of a slice. I believe like c8d0 instead off c8d0s1. I wanted to do this 1 disk at a time and let it resilver. what is the proper way to do this. I tried, I believe from memory : zpool replace -f rpool c8d1s1 c8d1 but it didn't let me do that. then I tried to put the disk offline first , but same result. Thanks, Dirk -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss