Re: [zfs-discuss] zfs-discuss Digest, Vol 33, Issue 19
Hi Neil, No problem, thanks for the info. I knew these cards were a gamble, but if I can get them working, it will be worth it. I've today spoken to the UK office of VMetro about drivers but I'm not holding out too much hope. They were very friendly and polite, but explained that they simply don't support these for end users, it's for large OEM's only. However, I did do a bit of digging this morning and found the e-mail address of Micro Memory's lead software developer, who looks to be the chap responsible for developing the Solaris drivers in the first place. So I'll be dropping him a line shortly and seeing if he can help at all. Ross 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] slog device
I'm not personally aware of any. The ioDrive from Fusion-io looks the most promising, but it's a new product from a new company so it's likely to be a while (if ever) before Solaris drivers appear. I've contacted them to ask about Solaris drivers, but haven't had a response yet. I summarised my findings in a long post about 2/3 of the way down this thread: http://www.opensolaris.org/jive/thread.jspa?threadID=65074tstart=30 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] slog device
PS. I note on the Fusion-io web page that they're working with HP to accelerate their servers. Would be nice if somebody from Sun could do the same (or let us know if Sun are working on similar technology). 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] confusion and frustration with zpool
On Tue, Jul 8, 2008 at 8:56 AM, Darren J Moffat [EMAIL PROTECTED] wrote: Pete Hartman wrote: I'm curious which enclosures you've had problems with? Mine are both Maxtor One Touch; the 750 is slightly different in that it has a FireWire port as well as USB. I've had VERY bad experiences with the Maxtor One Touch and ZFS. To the point that we gave up trying to use them. We last tried on snv_79 though. I've had bad experiences with the Seagate products. Last time I read a bunch of customer reviews on newegg.com and it seemed to be split between those with no issues and those with failures. My guess is that it's related to duty cycle - casual users who really don't beat up on the drive will have no problems, while power users will probably kill the drive. If my guess is correct, it's simply physics - lack of airflow over the HDA (head disk assembly). Regards, -- Al Hopper Logical Approach Inc,Plano,TX [EMAIL PROTECTED] Voice: 972.379.2133 Timezone: US CDT OpenSolaris Governing Board (OGB) Member - Apr 2005 to Mar 2007 http://www.opensolaris.org/os/community/ogb/ogb_2005-2007/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Thumper wedged somewhere in ZFS
Forwarding here, as suggested by chaps on storage-discuss. Just to clarify, I was running filebench directly on the x4500, not from an initiator, so this is probably not a COMSTAR thing. Ceri -- That must be wonderful! I don't understand it at all. -- Moliere ---BeginMessage--- We've got an x4500 running SXCE build 91 with stmf configured to share out a (currently) small number (9) of LUs to a (currently) small number of hosts (4). The x4500 is configured with ZFS root mirror, 6 RAIDZ sets across all six controllers, some hot spares in the gaps and a RAID10 set to use everything else up. Since this is an investigative setup, I have been running filebench locally on the x4500 to get some stats before moving on to do the same on the initiators against the x4500 and our current storage. While running the filebench OLTP workload with $filesize=5g on one of the RAIDZ pools, the x4500 seemed to hang while creating the fileset. On further investigation, a lot of things actually still worked; log in via SSH was fine, /usr/bin/ps worked ok, /usr/ucb/ps and any of the /usr/proc ptools just hung, man hung, and so on. savecore -L managed to do a dump but couldn't seem to exit. So I did a hard reset, the system came up fine and I actually do have the dump from savecore -L. I'm kind of out of my depth with mdb, but it looks pretty clear to me that all of the hung processes were somewhere in ZFS: # mdb -k unix.0 vmcore.0 mdb: failed to read panicbuf and panic_reg -- current register set will be unavailable Loading modules: [ unix genunix specfs dtrace cpu.generic cpu_ms.AuthenticAMD.15 uppc pcplusmp scsi_vhci zfs sd ip hook neti sctp arp usba fctl nca lofs md cpc random crypto nfs fcip logindmux ptm nsctl ufs sppp ipc ] ::memstat Page SummaryPagesMB %Tot Kernel3085149 12051 74% Anon20123780% Exec and libs3565130% Page cache 200779 7845% Free (cachelist) 193955 7575% Free (freelist)663990 2593 16% Total 4167561 16279 Physical 4167560 16279 ::pgrep ptree SPID PPID PGIDSIDUID FLAGS ADDR NAME R 1825 1820 1825 1803 0 0x4a004000 ff04f5096c80 ptree R 1798 1607 1798 1607 15000 0x4a004900 ff04f7b72930 ptree R 1795 1302 1795 1294 0 0x4a004900 ff05179f7de0 ptree ::pgrep ptree | ::walk thread | ::findstack stack pointer for thread ff04ea2ca440: ff00201777d0 [ ff00201777d0 _resume_from_idle+0xf1() ] ff0020177810 swtch+0x17f() ff00201778b0 turnstile_block+0x752() ff0020177920 rw_enter_sleep+0x1b0() ff00201779f0 zfs_getpage+0x10e() ff0020177aa0 fop_getpage+0x9f() ff0020177c60 segvn_fault+0x9ef() ff0020177d70 as_fault+0x5ae() ff0020177df0 pagefault+0x95() ff0020177f00 trap+0xbd3() ff0020177f10 0xfb8001d9() stack pointer for thread ff04e8752400: ff001f9307d0 [ ff001f9307d0 _resume_from_idle+0xf1() ] ff001f930810 swtch+0x17f() ff001f9308b0 turnstile_block+0x752() ff001f930920 rw_enter_sleep+0x1b0() ff001f9309f0 zfs_getpage+0x10e() ff001f930aa0 fop_getpage+0x9f() ff001f930c60 segvn_fault+0x9ef() ff001f930d70 as_fault+0x5ae() ff001f930df0 pagefault+0x95() ff001f930f00 trap+0xbd3() ff001f930f10 0xfb8001d9() stack pointer for thread ff066fbc6a80: ff001f27de90 [ ff001f27de90 _resume_from_idle+0xf1() ] ff001f27ded0 swtch+0x17f() ff001f27df00 cv_wait+0x61() ff001f27e040 vmem_xalloc+0x602() ff001f27e0b0 vmem_alloc+0x159() ff001f27e140 segkmem_xalloc+0x8c() ff001f27e1a0 segkmem_alloc_vn+0xcd() ff001f27e1d0 segkmem_zio_alloc+0x20() ff001f27e310 vmem_xalloc+0x4fc() ff001f27e380 vmem_alloc+0x159() ff001f27e410 kmem_slab_create+0x7d() ff001f27e450 kmem_slab_alloc+0x57() ff001f27e4b0 kmem_cache_alloc+0x136() ff001f27e4d0 zio_data_buf_alloc+0x28() ff001f27e510 arc_get_data_buf+0x175() ff001f27e560 arc_buf_alloc+0x9a() ff001f27e610 arc_read+0x122() ff001f27e6b0 dbuf_read_impl+0x129() ff001f27e710 dbuf_read+0xc5() ff001f27e7c0 dmu_buf_hold_array_by_dnode+0x1c4() ff001f27e860 dmu_read+0xd4() ff001f27e910 zfs_fillpage+0x15e() ff001f27e9f0 zfs_getpage+0x187() ff001f27eaa0 fop_getpage+0x9f() ff001f27ec60 segvn_fault+0x9ef() ff001f27ed70 as_fault+0x5ae() ff001f27edf0 pagefault+0x95() ff001f27ef00 trap+0xbd3() ff001f27ef10 0xfb8001d9() ::pgrep go_filebench | ::walk thread | ::findstack stack pointer for thread
Re: [zfs-discuss] slog device
Ross wrote: PS. I note on the Fusion-io web page that they're working with HP to accelerate their servers. Would be nice if somebody from Sun could do the same (or let us know if Sun are working on similar technology). I thought the cat was already out of the bag... :-) http://blogs.sun.com/jonathan/entry/not_a_flash_in_the -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs-discuss Digest, Vol 33, Issue 19
r == Ross [EMAIL PROTECTED] writes: np == Neil Perrin [EMAIL PROTECTED] writes: np 2. I received the board and driver from another group within np Sun. It would be better to contact Micro Memory (or whoever np took them over) directly, as it's not my place to give out 3rd np party drivers or provide support for them. Then hopefully when Sun releases their new batch of SSD devices, they will release source for the full driver stack under a redistributable license so that no well-meaning geek has to be in your awkwardly unhelpful position, caught between obligations of NDA/copyright/``place'' and the basic and reasonable obligations necessary to maintain a ``community''. I've heard Sun people at users' groups promise that all new Solaris subsystems will include source, but so far this doesn't apply to hardware, not even to the hardware Sun sells. In this case source would solve (1) and (2) because you'd be (2) free to redistribute whatever you had a month ago, and Ross would (1) have a fighting chance of forward-porting the driver he got from you. This isn't the case for existing Sun disk drivers that I know about like the X4500 SATA chip or the LSI Logic mpt RAID card in SPARC SATA systems, while Linux and I think BSD have free software drivers for both chips---at best the Sun drivers are (2) redistributable, and I'm not even clear on that because it's surprisingly tricky to determine. r they simply don't support these for end users, it's for large r OEM's only. [...] found the e-mail address of Micro Memory's r lead software developer, who, unlike the salespeople, will probably understand the obvious difference between providing ``support,'' and taking the basic responsibility to either archive all downloadables that aren't redistributable, or make them redistributable if they don't want to track them any more, but who probably won't be in a position to help you any more than Neil is. If their contractor did give you the drivers, would you avoid mentioning it here for fear a bunch of other people would ask you for copies, putting you in the same awkward spot? Would you justify the reticence by thinking you were hiding the drivers from us out of loyalty and ``gratitude'' to the contractor who wrote them? It stinks, and I recognize the smell. We've been here before. I ought to have better things to do with my life than pirating software to support obscure proprietary abandonware (but apparently not better than writing emails whining about the situation). pgpduxAm78jYK.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] confusion and frustration with zpool
ah == Al Hopper [EMAIL PROTECTED] writes: ah I've had bad experiences with the Seagate products. I've had bad experiences with all of them. (maxtor, hgst, seagate, wd) ah My guess is that it's related to duty cycle - Recently I've been getting a lot of drives from companies like newegg and zipzoomfly that fail within the first month. The rate is high enough that I would not trust a two-way mirror with 1mo old drives. Then I have drives with a few undreadable sectors 2 - 5 years into their life, from all manufacturers. I test them with 'smartctl -t long', and either send them for warranty repair or abandon them. I suspect usually 'dd if=/dev/zero of=drive' would fix such a disk unless the ``reallocated sector count'' is too high, but I just pretend every drive is on lease for its warranty period. The PATA/SATA/SATA2NCQ interfaces and capacity-per-watt changes about that often anyway. I send so many drives back for repair that it only makes financial sense to buy 5-year-warranty drives. I don't think they can make any money on me with the rate I send them, but if more people did this maybe they would learn to make disks that don't suck. Maybe they are giving me all their marginal ones or something, by using ``sales channels''---we pour our shit down THIS channel. In that case they could still make money. pgpsUTthi2URl.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Using zfs boot with MPxIO on T2000
Here is what I have configured: T2000 with OBP 4.28.6 2008/05/23 12:07 with 2 - 72 GB disks as the root disks OpenSolaris Nevada Build 91 Solaris Express Community Edition snv_91 SPARC Copyright 2008 Sun Microsystems, Inc. All Rights Reserved. Use is subject to license terms. Assembled 03 June 2008 Installed from DVD as ZFS boot filesystems 1 SAN disk attached from IBM SVC using DS8300 storage. 2 - 1 port Qlogic cards attached to McData 6064 Directors Here's my questions: 1. After the install I created a zfs mirror of the root disk c0t0d0 to c0t1d0, format shows the mirrored disk with sectors instead of cylinders, is this normal or correct? Is there a way to reverse this back to cylinders if it is not? Same goes for the external disk pool using SAN disk from the IBM SVC. 2. After enabled MPxIO (stmsboot -e), the 2 root disks now have MPxIO labels, is this a bug with ZFS boot using MPxIO? I have MPxIO running on Solaris 10 release 4 with none of this behavior (I have 2 T2000's, 1 with SVM root disks and other with Veritas Encapsulated root disks, all external or non root filesystems are managed by Veritas volume management, not ZFS). From format: 0. c4t5000C5000AF82EDBd0 SEAGATE-ST973402SSUN72G-0603-68.37GB /scsi_vhci/[EMAIL PROTECTED] 1. c4t5000C5000AF834ABd0 SUN72G cyl 14087 alt 2 hd 24 sec 424 /scsi_vhci/[EMAIL PROTECTED] 2. c4t60050768019081653617d0 IBM-2145--36.00GB /scsi_vhci/[EMAIL PROTECTED] root[:/root]# stmsboot -L non-STMS device nameSTMS device name -- /dev/rdsk/c0t0d0/dev/rdsk/c4t5000C5000AF834ABd0 /dev/rdsk/c0t1d0/dev/rdsk/c4t5000C5000AF82EDBd0 3. Any good references for using ZFS with MPxIO? Thanks in advance, Adrian 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] confusion and frustration with zpool
On Jul 9, 2008, at 11:12 AM, Miles Nordin wrote: ah == Al Hopper [EMAIL PROTECTED] writes: ah I've had bad experiences with the Seagate products. I've had bad experiences with all of them. (maxtor, hgst, seagate, wd) ah My guess is that it's related to duty cycle - Recently I've been getting a lot of drives from companies like newegg and zipzoomfly that fail within the first month. The rate is high enough that I would not trust a two-way mirror with 1mo old drives. While I've always had good luck with zipzoomfly, infant mortality is a well known feature of many devices. Your advice to do some burn in testing of drives before putting them into full production is probably a very sound one for sites large enough to maintain a bit of inventory ; -- Keith H. Bierman [EMAIL PROTECTED] | AIM kbiermank 5430 Nassau Circle East | Cherry Hills Village, CO 80113 | 303-997-2749 speaking for myself* Copyright 2008 ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs-discuss Digest, Vol 33, Issue 19
I think the problem Miles is that this isn't Sun hardware, and I completely understand that as a Sun employee, Neil really can't be seen to distribute something that's untested and unsupported, and quite possibly under NDA. On the other hand, if I get hold of these drivers, I'm under no such obligation and I'll be happily making them available for everybody who wants them. I already know of two other people who are keen to get these and I'm sure there are others. These cards are starting to show up on the second hand market now, finding a set of Solaris drivers would be a welcome bonus for a good few people. 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-discuss Digest, Vol 33, Issue 19
On Wed, Jul 9, 2008 at 1:14 PM, Ross [EMAIL PROTECTED] wrote: I think the problem Miles is that this isn't Sun hardware, and I completely understand that as a Sun employee, Neil really can't be seen to distribute something that's untested and unsupported, and quite possibly under NDA. On the other hand, if I get hold of these drivers, I'm under no such obligation and I'll be happily making them available for everybody who wants them. I already know of two other people who are keen to get these and I'm sure there are others. These cards are starting to show up on the second hand market now, finding a set of Solaris drivers would be a welcome bonus for a good few people. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss Do we have drivers available for ANY OS for these cards currently? It'd be nice to at least be able to test if they function properly. --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Using zfs boot with MPxIO on T2000
On Wed, Jul 9, 2008 at 6:27 PM, Adrian Danielson [EMAIL PROTECTED] wrote: Here is what I have configured: T2000 with OBP 4.28.6 2008/05/23 12:07 with 2 - 72 GB disks as the root disks OpenSolaris Nevada Build 91 ... 2. After enabled MPxIO (stmsboot -e), the 2 root disks now have MPxIO labels, is this a bug with ZFS boot using MPxIO? I have MPxIO running on Solaris 10 release 4 with none of this behavior (I have 2 T2000's, 1 with SVM root disks and other with Veritas Encapsulated root disks, all external or non root filesystems are managed by Veritas volume management, not ZFS). Nothing to do with ZFS. Current versions of the mpt driver, used in a lot of current Sun systems for the internal drive and for external SAS connectivity, support mpxio as well. (Solaris 10 update 4 doesn't have it - it came soon after in a patch.) You can restrict stmsboot to only enable mpxio on the mpt or fibre interfaces using 'stmsboot -D mpt' or 'stmsboot -D fp'. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] And the answer to why can't ZFS find a plugged in disk is ...
From an email exchange with a HAL developer... This comes about because I boot back and forth between Windows and Solaris and when on the Windows side I have the drive unplugged. On occasion, I forget to plug it back in before returning to Solaris. I wonder then, if Solaris should export removable ZFS volumes on shutdown. Seems a strange limitation for HAL to not attempt to mount a zfs file system. If it's not imported the mount fails and an error can be generated. If it's imported then everything just works. What was the reasoning for this? There are multiple reasons. Initially, when HAL was introduced in Solaris (PSARC 2005/399), ZFS did not support hotplug very well or at all. Also, HAL's object model only accomodates traditional single device volumes; it needs to be expanded to account for ZFS's volumes than span multiple devices. There are also more operations than just mount/unmount possible, and sometimes necessary, on ZFS datasets, and HAL simply lacks such interfaces. The third problematic area is that now that ZFS itself includes some sort of hotplug magic, there needs to be coordination with HAL-based volume managers. There are also potential difficulties related to different security models between traditionally mounted filesystems and ZFS. In other words, there is nothing fundamentally preventing HAL from supporting ZFS, but the amount of new design is enough for a full-blown project. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Using zfs boot with MPxIO on T2000
Adrian Danielson wrote: Here is what I have configured: T2000 with OBP 4.28.6 2008/05/23 12:07 with 2 - 72 GB disks as the root disks OpenSolaris Nevada Build 91 Solaris Express Community Edition snv_91 SPARC Copyright 2008 Sun Microsystems, Inc. All Rights Reserved. Use is subject to license terms. Assembled 03 June 2008 Installed from DVD as ZFS boot filesystems 1 SAN disk attached from IBM SVC using DS8300 storage. 2 - 1 port Qlogic cards attached to McData 6064 Directors Here's my questions: 1. After the install I created a zfs mirror of the root disk c0t0d0 to c0t1d0, format shows the mirrored disk with sectors instead of cylinders, is this normal or correct? Is there a way to reverse this back to cylinders if it is not? Same goes for the external disk pool using SAN disk from the IBM SVC. Please verify that you following the procedures for mirroring ZFS boot disks in the ZFS Adminstration Guide http://www.opensolaris.org/os/community/zfs/docs/zfsadmin.pdf As always, I also suggest testing prior to production roll-out. -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] RFE: ZFS commands zmv and zcp
I've run across something that would save me days of trouble. Situation, the contents of one ZFS file system needs to be moved to another ZFS file system. The destination can be the same Zpool, even a brand new ZFS file system. A command to move the data from one ZFS file system to another, WITHOUT COPYING, would be nice. At present, the data is almost 1TB. Ideally a zmv or zcp program would be nice. And no, zfs send and zfs receive won't do the same thing. Those would require hours, or possibly days to copy 1TB to the same Zpool. Plus, I would have to make the source R/O during the copy. I can create a new Zpool or send the data to another Zpool that has space, but then I end up with a 1TB of un-used space on the original Zpool. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] previously mentioned J4000 released
Here's the announcement for those new Sun JBOD devices mentioned the other day. http://www.sun.com/aboutsun/pr/2008-07/sunflash.20080709.1.xml ckl ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] RFE: ZFS commands zmv and zcp
Raquel K. Sanborn wrote: I've run across something that would save me days of trouble. Situation, the contents of one ZFS file system needs to be moved to another ZFS file system. The destination can be the same Zpool, even a brand new ZFS file system. A command to move the data from one ZFS file system to another, WITHOUT COPYING, would be nice. At present, the data is almost 1TB. Ideally a zmv or zcp program would be nice. And no, zfs send and zfs receive won't do the same thing. Those would require hours, or possibly days to copy 1TB to the same Zpool. Plus, I would have to make the source R/O during the copy. I can create a new Zpool or send the data to another Zpool that has space, but then I end up with a 1TB of un-used space on the original Zpool. Please follow the thread discussed here last December. http://mail.opensolaris.org/pipermail/zfs-discuss/2007-December/044975.html -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] RFE: ZFS commands zmv and zcp
On Wed, 9 Jul 2008, Raquel K. Sanborn wrote: Situation, the contents of one ZFS file system needs to be moved to another ZFS file system. The destination can be the same Zpool, even a brand new ZFS file system. A command to move the data from one ZFS file system to another, WITHOUT COPYING, would be nice. At present, the data is almost 1TB. I agree that this would be quite useful. Is it possible that snapshot + clone + promote could be useful for your current purpose? 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] X4540
So, I see Sun finally updated the Thumper, and it appears they're now using a PCI-E backplane. Anyone happen to know what the chipset is? Any chance we'll see an 8-port PCI-E SATA card finally?? The new Sun Fire X4540 server uses PCI Express IO technology for more than triple the system IO-to-network bandwidth. http://www.sun.com/servers/x64/x4540/ --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] previously mentioned J4000 released
On Wed, Jul 9, 2008 at 2:39 PM, Chad Lewis [EMAIL PROTECTED] wrote: Here's the announcement for those new Sun JBOD devices mentioned the other day. http://www.sun.com/aboutsun/pr/2008-07/sunflash.20080709.1.xml ckl ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss So are these *tagged* drives/firmware? Do we have to buy them direct from Sun or can we throw anything we want at it? Does it come pre-loaded with real drive trays instead of useless blanks? --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] X4540
The X4540 uses on-board LSI SAS controllers (C1068E). - Eric On Wed, Jul 09, 2008 at 02:59:26PM -0500, Tim wrote: So, I see Sun finally updated the Thumper, and it appears they're now using a PCI-E backplane. Anyone happen to know what the chipset is? Any chance we'll see an 8-port PCI-E SATA card finally?? The new Sun Fire X4540 server uses PCI Express IO technology for more than triple the system IO-to-network bandwidth. http://www.sun.com/servers/x64/x4540/ --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Eric Schrock, Fishworkshttp://blogs.sun.com/eschrock ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] X4540
On Wed, Jul 9, 2008 at 3:09 PM, Eric Schrock [EMAIL PROTECTED] wrote: The X4540 uses on-board LSI SAS controllers (C1068E). - Eric On Wed, Jul 09, 2008 at 02:59:26PM -0500, Tim wrote: So, I see Sun finally updated the Thumper, and it appears they're now using a PCI-E backplane. Anyone happen to know what the chipset is? Any chance we'll see an 8-port PCI-E SATA card finally?? The new Sun Fire X4540 server uses PCI Express IO technology for more than triple the system IO-to-network bandwidth. http://www.sun.com/servers/x64/x4540/ --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Eric Schrock, Fishworks http://blogs.sun.com/eschrock Perfect. Which means good ol' supermicro would come through :) WOHOO! AOC-USAS-L8i http://www.supermicro.com/products/accessories/addon/AOC-USAS-L8i.cfm --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs-discuss Digest, Vol 33, Issue 19
r == Ross [EMAIL PROTECTED] writes: r I think the problem Miles is that this isn't Sun hardware In this case it's not, but please do not muddle my point: Marvell SATA and LSI Logic mpt SATARAID and many other (most?) drivers have the same problem. Right now there are, AIUI: * closed-source non-redistributable drivers (SXCE only) * closed-source redistributable drivers (SXCE, Indiana, Nexenta) * open-source redistributable drivers (SXCE, Indiana, Nexenta) The logical fourth category of open-source non-redistributable drivers doesn't exist---you CAN have a non-redistributable, $0 driver which includes source code, but it wouldn't meet the open source specification. The word ``third-party driver'' is thrown around a lot. I guess it was a common word in the pre-Opensolaris days? The three categories are orthogonal to bundling or support entitlements, and there are plenty of Solaris/SXCE-bundled, support-entitled drivers in the first category. r I completely understand that as a Sun employee, Neil really r can't be seen to distribute something that's untested and r unsupported, and quite possibly under NDA. AIUI it's not personal, or be-seen-as. It's, do you have the right to do it, or do you not. For example, I do not have the right to give you an SXCE DVD. You have to download it yourself. (hope you don't need an old version!) I DO have the right to give you a Nexenta or OpenSolaris 2008.05 DVD. This is redistribution. To pass redistribution rights on to me, Sun left drivers out of the OpenSolaris/Indiana release and Nexenta out of the Nexenta release. And just as Micro Memory can take a formerly-$0 driver down from their web page, Sun can take down the SXCE b12345 .iso, and if you don't already have a copy hoarded you're not technically allowed to have your friend copy his DVD and give it to you. r if I get hold of these drivers, I'm under no such obligation The obligation would come when you get the drivers---you'll be given drivers on the condition you agree to something. Since you don't have them yet, you're in a bad position to promise this. You could promise, ``I won't accept drivers from anyone who makes me promise not to redistribute them or not to release the source code of them,'' (or publish benchmarks without the manufacturers approval COUGH COUGH) which is what I _wish_ Sun would do to the chip and card vendors from which they get components in the hardware they sell, but they don't. You could also promise, ``If someone makes me agree not to redistribute this, I'll agree and then break the agreement, because I care more about preserving the community than I do about respecting legal agreements.'' To me it seems like technical people take exclusively the former approach, and casual non-technical users almost exclusively the latter. I guess there are a lot of people in the world who can repeatedly make the latter statement publicly without hurting their careers, but maybe not many such people on this list. anyway sorry it's OT. I'll drop it now. I should hunt for a [EMAIL PROTECTED] list or something. pgpzb16u6QFN2.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] X4540
On Wed, Jul 9, 2008 at 2:59 PM, Tim [EMAIL PROTECTED] wrote: So, I see Sun finally updated the Thumper, and it appears they're now using a PCI-E backplane. Anyone happen to know what the chipset is? Any chance we'll see an 8-port PCI-E SATA card finally?? The new Sun Fire X4540 server uses PCI Express IO technology for more than triple the system IO-to-network bandwidth. http://www.sun.com/servers/x64/x4540/ Any word on why PCI-Express was not extended to the expansion slots? I put PCI-Express cards in every other server that I connect to 10 gigabit Ethernet or the SAN (FC tape drives). -- Mike Gerdts http://mgerdts.blogspot.com/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] previously mentioned J4000 released
On Wed, Jul 9, 2008 at 2:39 PM, Chad Lewis [EMAIL PROTECTED] wrote: Here's the announcement for those new Sun JBOD devices mentioned the other day. http://www.sun.com/aboutsun/pr/2008-07/sunflash.20080709.1.xml ckl ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss Very interesting, I have two questions: Does this require tagged drives? IE: do we *HAVE* to purchase all drives that go into these direct from Sun? Does it ship with real drive trays in the *empty* slots, or those worthless blanks that won't hold a drive? --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] X4540
Tim wrote: So, I see Sun finally updated the Thumper, and it appears they're now using a PCI-E backplane. Anyone happen to know what the chipset is? Any chance we'll see an 8-port PCI-E SATA card finally?? One NVidia MCP-55 and two NVidia IO-55s replace the thumper's AMD-8132 HT to PCI-X bridges. The new configuration is such that the expandable PCI-E slots have their own IO-55. The MCP-55 and one IO-55 connect to 3 LSI 1068E and provide 2x GbE each. This should be a better balance than the thumper's configuration. LSI 1068E SAS/SATA controllers replace thumper's Marvell SAS/SATA controllers. You might recognize the LSI 1068, and its smaller cousin, the 1064, as being used in many other Sun servers from the T1000 to the M9000. 8-port PCI-E SAS/SATA card is supported for additional expansion, such as a J4500 (the JBOD-only version) http://www.sun.com/storagetek/storage_networking/hba/sas/specs.xml The best news, for many folks, is that you can boot from an (externally pluggable) CF card, so that you don't have to burn two disks for the OS. I think we have solved many of the deficiencies noted in the thumper, including more CPU and memory capacity. Please let us know what you think :-) -- richard The new Sun Fire X4540 server uses PCI Express IO technology for more than triple the system IO-to-network bandwidth. http://www.sun.com/servers/x64/x4540/ --Tim ___ 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] X4540
On Wed, Jul 09, 2008 at 03:19:53PM -0500, Mike Gerdts wrote: Any word on why PCI-Express was not extended to the expansion slots? I put PCI-Express cards in every other server that I connect to 10 gigabit Ethernet or the SAN (FC tape drives). The webpage is incorrect. There are three 8x PCI-E half-height slots on the X4540. - Eric -- Eric Schrock, Fishworkshttp://blogs.sun.com/eschrock ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] RFE: ZFS commands zmv and zcp
Thanks, glad someone else thought of it first. I guess I will have to do things the hard way. Raquel 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] X4540
Might also want to have them talk to byteandswitch. * **We went to the next-generation Intel processors [and] we have used the latest generation of our Solaris ZFS software, he explains, adding that the J4000 JBODs can also be connected to the X4540.* Either the 4540 is using XEON's now, someone was misquoted, or someone was confused :) http://www.byteandswitch.com/document.asp?doc_id=158533WT.svl=news1_1 --Tim On Wed, Jul 9, 2008 at 3:44 PM, Richard Elling [EMAIL PROTECTED] wrote: Yes, thanks for catching this. I'm sure it is just a copy-n-paste mistake. I've alerted product manager to get it fixed. -- richard Mike Gerdts wrote: On Wed, Jul 9, 2008 at 3:29 PM, Richard Elling [EMAIL PROTECTED] wrote: 8-port PCI-E SAS/SATA card is supported for additional expansion, such as a J4500 (the JBOD-only version) http://www.sun.com/storagetek/storage_networking/hba/sas/specs.xml Based upon my previous message, this message, and Jeorg Moellenkamp's blog entry[1], I think that the hardware specifications page[2] needs to be updated so that the expansion slots say PCI-Express rather than PCI-X. 1. http://www.c0t0d0s0.org/archives/4605-New-storage-from-Sun-J420044004500-and-X4540-Storage-Server.html 2. http://www.sun.com/servers/x64/x4540/specs.xml ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] X4540
On Wed, Jul 09, 2008 at 03:52:27PM -0500, Tim wrote: Is the 4540 still running a rageXL? I find that somewhat humorous if it's an Nvidia chipset with ATI video :) According to SMBIOS there is an on-board device of type AST2000 VGA. - Eric -- Eric Schrock, Fishworkshttp://blogs.sun.com/eschrock ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] X4540
Tim wrote: Is the 4540 still running a rageXL? I find that somewhat humorous if it's an Nvidia chipset with ATI video :) Yes, it is part of the chip which handles the management interface. I don't find this to be a contradiction, though. AMD bought ATI and we're using AMD Quad-core CPUs. -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Using zfs boot with MPxIO on T2000
Adrian Danielson wrote: 1. After the install I created a zfs mirror of the root disk c0t0d0 to c0t1d0, format shows the mirrored disk with sectors instead of cylinders, is this normal or correct? Is there a way to reverse this back to cylinders if it is not? Same goes for the external disk pool using SAN disk from the IBM SVC. Format show sectors when the disk has an EFI label, and cylinders when the disk has a Sun label. ZFS always uses EFI labels, so you're seeing the right thing. You can change the label (blowing away the disk contents of course) using format -e. The label menu changes with the -e flag to let you choose the kind of label. Cheers, Chris ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Using zfs boot with MPxIO on T2000
ZFS uses EFI when a storage pool is created with whole disks. ZFS uses the old-style VTOC label when a storage pool is created with slices. To be able to boot from a ZFS root pool, the storage pool must be created with slices. This is a new requirement in ZFS land, and is described in the doc pointer Richard provided previously: http://www.opensolaris.org/os/community/zfs/docs/zfsadmin.pdf Cindy Chris Ridd wrote: Adrian Danielson wrote: 1. After the install I created a zfs mirror of the root disk c0t0d0 to c0t1d0, format shows the mirrored disk with sectors instead of cylinders, is this normal or correct? Is there a way to reverse this back to cylinders if it is not? Same goes for the external disk pool using SAN disk from the IBM SVC. Format show sectors when the disk has an EFI label, and cylinders when the disk has a Sun label. ZFS always uses EFI labels, so you're seeing the right thing. You can change the label (blowing away the disk contents of course) using format -e. The label menu changes with the -e flag to let you choose the kind of label. Cheers, Chris ___ 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] Using zfs boot with MPxIO on T2000
Peter Tribble wrote: On Wed, Jul 9, 2008 at 6:27 PM, Adrian Danielson [EMAIL PROTECTED] wrote: Here is what I have configured: T2000 with OBP 4.28.6 2008/05/23 12:07 with 2 - 72 GB disks as the root disks OpenSolaris Nevada Build 91 ... 2. After enabled MPxIO (stmsboot -e), the 2 root disks now have MPxIO labels, is this a bug with ZFS boot using MPxIO? I have MPxIO running on Solaris 10 release 4 with none of this behavior (I have 2 T2000's, 1 with SVM root disks and other with Veritas Encapsulated root disks, all external or non root filesystems are managed by Veritas volume management, not ZFS). Nothing to do with ZFS. Current versions of the mpt driver, used in a lot of current Sun systems for the internal drive and for external SAS connectivity, support mpxio as well. (Solaris 10 update 4 doesn't have it - it came soon after in a patch.) You can restrict stmsboot to only enable mpxio on the mpt or fibre interfaces using 'stmsboot -D mpt' or 'stmsboot -D fp'. Hi Adrian, as Peter mentions, this isn't a bug, it's a feature ;) Actually, it's the feature that I delivered into Solaris 10 last year with the 125081-10/125082-10 patches. 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] X4540
Eric Schrock wrote: On Wed, Jul 09, 2008 at 03:52:27PM -0500, Tim wrote: Is the 4540 still running a rageXL? I find that somewhat humorous if it's an Nvidia chipset with ATI video :) According to SMBIOS there is an on-board device of type AST2000 VGA. Yes, I think I found another copy-n-paste error in some docs :-( It does appear to be an AST2000, something like: http://www.aspeedtech.com/ast2000.html -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] X4540
On Wed, Jul 9, 2008 at 3:29 PM, Richard Elling [EMAIL PROTECTED] wrote: Tim wrote: So, I see Sun finally updated the Thumper, and it appears they're now using a PCI-E backplane. Anyone happen to know what the chipset is? Any chance we'll see an 8-port PCI-E SATA card finally?? One NVidia MCP-55 and two NVidia IO-55s replace the thumper's AMD-8132 HT to PCI-X bridges. The new configuration is such that the expandable PCI-E slots have their own IO-55. The MCP-55 and one IO-55 connect to 3 LSI 1068E and provide 2x GbE each. This should be a better balance than the thumper's configuration. LSI 1068E SAS/SATA controllers replace thumper's Marvell SAS/SATA controllers. You might recognize the LSI 1068, and its smaller cousin, the 1064, as being used in many other Sun servers from the T1000 to the M9000. 8-port PCI-E SAS/SATA card is supported for additional expansion, such as a J4500 (the JBOD-only version) http://www.sun.com/storagetek/storage_networking/hba/sas/specs.xml The best news, for many folks, is that you can boot from an (externally pluggable) CF card, so that you don't have to burn two disks for the OS. I think we have solved many of the deficiencies noted in the thumper, including more CPU and memory capacity. Please let us know what you think :-) Not that I'm in the market for one - but I think a version with (possibly fewer) 15k RPM SAS disks would be a best seller - especially for applications that require more IOPS. Like RDBMS for example. And yes, I realize that one could install a SAS card into the 4540 and attach it to one of the SAS based J4nnn boxes - but that's not the same physical density that a 4540 with SAS disks would offer. Or even a mixture of SATA and SAS drives And it would be great if Sun would OEM the Micro Memory (aka vmetro) cards. Obviously its only a question of time before Sun will bring its own RAM/flash cards to the market - but an OEM deal would make product available now and probably won't compete with what Sun has in mind (based entirely on my own crystal ball gazing). We all know how big a win this is for NFS shares! Congrats to Sun, Team ZFS and open storage. The new x45xx and J4xxx boxes are *great* additions to Suns product line. -- richard The new Sun Fire X4540 server uses PCI Express IO technology for more than triple the system IO-to-network bandwidth. http://www.sun.com/servers/x64/x4540/ --Tim Regards, -- Al Hopper Logical Approach Inc,Plano,TX [EMAIL PROTECTED] Voice: 972.379.2133 Timezone: US CDT OpenSolaris Governing Board (OGB) Member - Apr 2005 to Mar 2007 http://www.opensolaris.org/os/community/ogb/ogb_2005-2007/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Case study/recommended ZFS setup for home file server
Hello, I plan to use (Open)Solaris for a home file server. I wanted cool and quiet hardware, so I picked a mini-atx motherboard and case, an AMD64 CPU and 4 GB of RAM. My case has room for three hard drives and I have chosen 3x WD 750 Green Power hard drives. The file server will serve out via NFS and Samba the home directories, the library (collected articles and books in PDF format) and the photo archive (150GB and growing of photos in RAW format ~ 7-9MB/file). I cannot use OpenSolaris 2008.05 since it does not recognize the SATA disks attached to the southbridge. A fix for this problem went into build 93. I will use SXCE 93 (for the SATA fix) or SXCE 94 (for the last revision of the ZFS format). In order to make the maximum amount of space available for the photos, I plan to use RAID-5 for that pool. Also, I would like to have sufficient redundancy so if a drive goes bad, I can just replace it and the volume manger/file system will take care of fixing itself back. The question is, how should I partition the drives, and what tuning parameters should I use for the pools and file systems? From reading the best practices guides [1], [2], it seems that I cannot have the root file system on a RAID-5 pool, but it has to be a separate storage pool. This seems to be slightly at odds with the suggestion of using whole-disks for ZFS, not just slices/partitions. My plan right now is to create a 20 GB and a 720 GB slice on each disk, then create two storage pools, one RAID-1 (20 GB) and one RAID-5 (1.440 TB). Create the root, var, usr and opt file systems in the first pool, and home, library and photos in the second. I hope I won't need swap, but I could create three 1 GB slices (one on each disk) for that. Does this sound like a good configuration? Will the SXCE 9[34] installer allow me to create the above setup? Should I pass any special parameters to the zfs pool and file system creation tool to get the best performance? home and library contains files between few KB and a fer MB. photos contains file roughly 7 to 9 MB. Should I place those on separate pools? Note: the hardware is committed (i.e. I already have it), so I am not inclined to deviate from it 8^) Thanks, florin 1: http://www.solarisinternals.com/wiki/index.php/ZFS_Best_Practices_Guide 2: http://www.solarisinternals.com/wiki/index.php/ZFS_Configuration_Guide -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 pgp9qLioSeY7W.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] slog device
On Tue, Jul 8, 2008 at 10:42 PM, Gilberto Mautner [EMAIL PROTECTED] wrote: Hi, Anyway, are there other devices out there that you would recommend to use as a slog device, other than this nvram card, that would present similar performance gains? Not that this will get you similar performance gains - but don't overlook putting a couple of small 15k RPM SAS disk drives in the box. They work great with ZFS and really help out those poor SATA drives when ZFS starts beating up on them.And it also helps if you can't put more RAM in the box. Conduct your own experiments with 15k SAS drives as slog/cache devices. Worst case scenario, you'll simply end up using them as ZFS vdevs. :) Regards, -- Al Hopper Logical Approach Inc,Plano,TX [EMAIL PROTECTED] Voice: 972.379.2133 Timezone: US CDT OpenSolaris Governing Board (OGB) Member - Apr 2005 to Mar 2007 http://www.opensolaris.org/os/community/ogb/ogb_2005-2007/ ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] confusion and frustration with zpool
Also worth noting is that the enterprise-class drives have protection from heavy load that the consumer-class drives don't. In particular, there's no temperature sensor on the voice coil for the consumer drives, which means that under heavy seek load (constant i/o), the drive will eventually overheat. [There are plenty of other differences, but this one is important if you plan to put a drive into 24/7 use.] This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Supermicro AOC-USAS-L8i
On Wed, Jul 9, 2008 at 1:12 PM, Tim [EMAIL PROTECTED] wrote: Perfect. Which means good ol' supermicro would come through :) WOHOO! AOC-USAS-L8i http://www.supermicro.com/products/accessories/addon/AOC-USAS-L8i.cfm Is this card new? I'm not finding it at the usual places like Newegg, etc. It looks like the LSI SAS3081E-R, but probably at 1/2 the cost. -B -- Brandon High [EMAIL PROTECTED] The good is the enemy of the best. - Nietzsche ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Supermicro AOC-USAS-L8i
On Thu, Jul 10, 2008 at 10:34 AM, Brandon High [EMAIL PROTECTED] wrote: On Wed, Jul 9, 2008 at 1:12 PM, Tim [EMAIL PROTECTED] wrote: Perfect. Which means good ol' supermicro would come through :) WOHOO! AOC-USAS-L8i http://www.supermicro.com/products/accessories/addon/AOC-USAS-L8i.cfm Is this card new? I'm not finding it at the usual places like Newegg, etc. It looks like the LSI SAS3081E-R, but probably at 1/2 the cost. It appears to be the non--RAID version of the card. (That's what the R suffix indicates). If it is that is the case, then I've got one running quite happily in my workstation already, using the mpt driver. James C. McPherson -- Solaris kernel software engineer, system admin and troubleshooter http://www.jmcp.homeunix.com/blog http://blogs.sun.com/jmcp Find me on LinkedIn @ http://www.linkedin.com/in/jamescmcpherson ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Supermicro AOC-USAS-L8i
Dunno how old it is, but James is right, no Raid which is why it's cheaper. Also why I like it ;) On Wed, Jul 9, 2008 at 7:34 PM, Brandon High [EMAIL PROTECTED] wrote: On Wed, Jul 9, 2008 at 1:12 PM, Tim [EMAIL PROTECTED] wrote: Perfect. Which means good ol' supermicro would come through :) WOHOO! AOC-USAS-L8i http://www.supermicro.com/products/accessories/addon/AOC-USAS-L8i.cfm Is this card new? I'm not finding it at the usual places like Newegg, etc. It looks like the LSI SAS3081E-R, but probably at 1/2 the cost. -B -- Brandon High [EMAIL PROTECTED] The good is the enemy of the best. - Nietzsche ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Case study/recommended ZFS setup for home file server
On Wed, Jul 9, 2008 at 3:37 PM, Florin Iucha [EMAIL PROTECTED] wrote: The question is, how should I partition the drives, and what tuning parameters should I use for the pools and file systems? From reading the best practices guides [1], [2], it seems that I cannot have the root file system on a RAID-5 pool, but it has to be a separate storage pool. This seems to be slightly at odds with the suggestion of using whole-disks for ZFS, not just slices/partitions. The reason for using a whole disk is that ZFS will turn on the drive's cache. When using slices, the cache is normally disabled. If all slices are using ZFS, you can turn the drive cache back on. I don't think it happens by default right now, but you can set it manually. Another alternative is to use an IDE to Compact Flash adapter, and boot off of flash. I'll be building a media server once we move, and that system will boot from flash. You can also boot from USB keys, but USB under OpenSolaris seems to be iffy. Here's the component list that I'm planning to use right now: http://secure.newegg.com/WishList/PublicWishDetail.aspx?Source=MSWDWishListNumber=7739092 I *may* change it and boot off another drive that is not part of the RAID-Z pool. My plan right now is to create a 20 GB and a 720 GB slice on each disk, then create two storage pools, one RAID-1 (20 GB) and one RAID-5 (1.440 TB). Create the root, var, usr and opt file systems in the first pool, and home, library and photos in the second. I hope I won't need swap, but I could create three 1 GB slices (one on each disk) for that. Does this sound like a good configuration? If you have enough memory (say 4gb) you probably won't need swap. I believe swap can live in a ZFS pool now too, so you won't necesarily need another slice. You'll just have RAID-Z protected swap. I built a Linux-based NAS a few years back using an almost identical scheme and wound up regretting it. In the future I would install the system on a completely separate disk or group of disks than the shared pool. Should I pass any special parameters to the zfs pool and file system creation tool to get the best performance? home and library contains files between few KB and a fer MB. photos contains file roughly 7 to 9 MB. Should I place those on separate pools? You shouldn't need to do anything. If you want to set the block size, or enable or disable compression, etc. you can create multiple filesystems in your pool rather than multiple pools. Note: the hardware is committed (i.e. I already have it), so I am not inclined to deviate from it 8^) You might want to look at a 4 or 8 port SATA adapter rather than wait for the southbridge fixes. -B -- Brandon High [EMAIL PROTECTED] The good is the enemy of the best. - Nietzsche ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Supermicro AOC-USAS-L8i
On Wed, Jul 09, 2008 at 07:53:30PM -0500, Tim wrote: Dunno how old it is, but James is right, no Raid which is why it's cheaper. Also why I like it ;) I have the HP badged LSA SAS3080X in my Ultra80, it's a fantastic card. If I ever get a box with PCI-E (I'm looking to upgrade the U80 soon, so it might just happen) that card looks like it would be *perfect*. On Wed, Jul 9, 2008 at 7:34 PM, Brandon High [EMAIL PROTECTED] wrote: On Wed, Jul 9, 2008 at 1:12 PM, Tim [EMAIL PROTECTED] wrote: Perfect. Which means good ol' supermicro would come through :) WOHOO! AOC-USAS-L8i http://www.supermicro.com/products/accessories/addon/AOC-USAS-L8i.cfm Is this card new? I'm not finding it at the usual places like Newegg, etc. It looks like the LSI SAS3081E-R, but probably at 1/2 the cost. -B -- Brandon High [EMAIL PROTECTED] The good is the enemy of the best. - Nietzsche ___ 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] Case study/recommended ZFS setup for home file
Thank you for starting this thread. I hope you get some good feedback. Your questions are quite frequently asked in this forum, but I'm very interested in the topic. Anway, the best answer varies from month to month. So I hope you get some good feedback. I cannot use OpenSolaris 2008.05 since it does not recognize the SATA disks attached to the southbridge. A fix for this problem went into build 93. Which forum/mailing list discusses SATA issues like the above? Thanks 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] Case study/recommended ZFS setup for home file
On Wed, Jul 09, 2008 at 08:42:37PM -0700, Bohdan Tashchuk wrote: I cannot use OpenSolaris 2008.05 since it does not recognize the SATA disks attached to the southbridge. A fix for this problem went into build 93. Which forum/mailing list discusses SATA issues like the above? #opensolaris in freenode.net I booted from the OpenSolaris LiveCD/installer, and noticing the lack of available disks, I cried for help on #irc. There were a few helpful people that gave me some commands to run, to try and get this going. After their efforts failed, I googled for solaris and SB600 (this is the ATI SouthBridge chip) and found a forum posting from another user, back in February, and the hit in bugzilla, pointing to the resolution of the bug, with the target being snv_93. Cheers, florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 pgpBfvJz8y2GS.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Case study/recommended ZFS setup for home file server
On Wed, Jul 09, 2008 at 06:02:24PM -0700, Brandon High wrote: On Wed, Jul 9, 2008 at 3:37 PM, Florin Iucha [EMAIL PROTECTED] wrote: The reason for using a whole disk is that ZFS will turn on the drive's cache. When using slices, the cache is normally disabled. If all slices are using ZFS, you can turn the drive cache back on. I don't think it happens by default right now, but you can set it manually. Aha! Good to know. Another alternative is to use an IDE to Compact Flash adapter, and boot off of flash. I'll be building a media server once we move, and that system will boot from flash. You can also boot from USB keys, but USB under OpenSolaris seems to be iffy. Here's the component list that I'm planning to use right now: http://secure.newegg.com/WishList/PublicWishDetail.aspx?Source=MSWDWishListNumber=7739092 That adapter won't work for me, since I have a single IDE port, and I need to use the DVD to install the OS and maybe to run some backups. However, this looks interesting: http://www.addonics.com/products/flash_memory_reader/ad2sahdcf.asp as it has hardware mirroring. Not sure what the error reporting through the OS is though..., but I hope I don't have to find out. For the Compact Flash I would spring for the industrial grade: http://www.hitechvendors.com/showproduct.aspx?ProductID=4885SEName=transcend-4gb-100x-industrial-cf-card-udma4-mode My plan right now is to create a 20 GB and a 720 GB slice on each disk, then create two storage pools, one RAID-1 (20 GB) and one RAID-5 (1.440 TB). Create the root, var, usr and opt file systems in the first pool, and home, library and photos in the second. I hope I won't need swap, but I could create three 1 GB slices (one on each disk) for that. I built a Linux-based NAS a few years back using an almost identical scheme and wound up regretting it. In the future I would install the system on a completely separate disk or group of disks than the shared pool. This is the current Linux-based NAS and I'm not happy with its performance, either. Note: the hardware is committed (i.e. I already have it), so I am not inclined to deviate from it 8^) You might want to look at a 4 or 8 port SATA adapter rather than wait for the southbridge fixes. I like the southbridge since it sits on the PCI express bus. The PCI bus is limited to 133 MB/s, which divided by 3 (disks) means 35-40 MB/s (including overhead) writes. And good quality PCI-express add-on controllers with Solaris drivers are quite expensive. Cheers, florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 pgpWxnTi94UBz.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Using zfs boot with MPxIO on T2000
Thanks everyone for your replies, I have a better understanding of how to test out ZFS with MPxIO. Best Regards, Adrian 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] Case study/recommended ZFS setup for home file
Florin Iucha wrote: On Wed, Jul 09, 2008 at 08:42:37PM -0700, Bohdan Tashchuk wrote: I cannot use OpenSolaris 2008.05 since it does not recognize the SATA disks attached to the southbridge. A fix for this problem went into build 93. Which forum/mailing list discusses SATA issues like the above? #opensolaris in freenode.net I booted from the OpenSolaris LiveCD/installer, and noticing the lack of available disks, I cried for help on #irc. There were a few helpful people that gave me some commands to run, to try and get this going. After their efforts failed, I googled for solaris and SB600 (this is the ATI SouthBridge chip) and found a forum posting from another user, back in February, and the hit in bugzilla, pointing to the resolution of the bug, with the target being snv_93. The OpenSolaris live CD has a hardware device detection tool. Please run it and submit the results (everyone should do this :-) b93 should be out soon, in the next week or so. -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Using zfs boot with MPxIO on T2000
http://www.opensolaris.org/os/community/zfs/docs/zfsadmin.pdf Another question: When trying to mirror the 2nd root disk, I get the following error: root[tst01:/root]# zpool attach rpool c0t0d0s0 c0t1d0s0 cannot attach c0t1d0s0 to c0t0d0s0: device is too small Since c0t1d0s0 has been reduced by 1 cylinder it's too small, is there a way to reduce the existing rpool so it will fit? I did not see in the zfsadmin.pdf guide if there was a way to do this or a work around. If I use the -f it will work but create an EFI labeled disk as I understand does not boot using ZFS. I must be overlooking a step. Thanks again, Adrian 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] Using zfs boot with MPxIO on T2000
Adrian Danielson wrote: http://www.opensolaris.org/os/community/zfs/docs/zfsadmin.pdf Another question: When trying to mirror the 2nd root disk, I get the following error: root[tst01:/root]# zpool attach rpool c0t0d0s0 c0t1d0s0 cannot attach c0t1d0s0 to c0t0d0s0: device is too small Since c0t1d0s0 has been reduced by 1 cylinder it's too small, is there a way to reduce the existing rpool so it will fit? I did not see in the zfsadmin.pdf guide if there was a way to do this or a work around. If I use the -f it will work but create an EFI labeled disk as I understand does not boot using ZFS. I must be overlooking a step. Sorry, you're out of luck - at least for the moment. Can you create the rpool to be smaller? 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