Re: [zfs-discuss] zfs code and fishworks fork
Hi all, I fully understand that within a cost effective point of view, developing the fishworks for a reduced set of hardware makes , alot, of sense. However, i think that Sun/Oracle would increase their user base if they make availabe a Fishwork framework certified only for a reduced set of hardware, ie : * it needs Western Digital HDD firmware version x.y.z * it needs a SAS/SATA controller from a specific brand, model and firmare ( LSI SAS1068E ) * if SSD's are used they need to be from vendor X with firmware Y * the system motherboard chipset needs to be from vendor X or Y and not from Z Within this possible landscape i'm pretty sure that alot more customers would pay for the Fishworks stack and support, given the fact that not all customers need aKa can afford, the Unified Storage platform from Sun. Anyway..Fishworks it's an awesome product! Congratulations for the extreme good job. Regards, Bruno Adam Leventhal wrote: With that said I'm concerned that there appears to be a fork between the opensource version of ZFS and ZFS that is part of the Sun/Oracle FishWorks 7nnn series appliances. I understand (implicitly) that Sun (/Oracle) as a commercial concern, is free to choose their own priorities in terms of how they use their own IP (Intellectual Property) - in this case, the source for the ZFS filesystem. Hey Al, I'm unaware of specific plans for management either at Sun or at Oracle, but from an engineering perspective suffice it to say that it is simpler and therefore more cost effective to develop for a single, unified code base, to amortize the cost of testing those modifications, and to leverage the enthusiastic ZFS community to assist with the development and testing of ZFS. Again, this isn't official policy, just the simple facts on the ground from engineering. I'm not sure what would lead you to believe that there is fork between the open source / OpenSolaris ZFS and what we have in Fishworks. Indeed, we've made efforts to make sure there is a single ZFS for the reason stated above. Any differences that exist are quickly migrated to ON as you can see from the consistent work of Eric Schrock. Adam -- Adam Leventhal, Fishworkshttp://blogs.sun.com/ahl ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] root pool can not have multiple vdevs ?
This seems like a bit of a restriction ... is this intended ? # cat /etc/release Solaris Express Community Edition snv_125 SPARC Copyright 2009 Sun Microsystems, Inc. All Rights Reserved. Use is subject to license terms. Assembled 05 October 2009 # uname -a SunOS neptune 5.11 snv_125 sun4u sparc SUNW,Sun-Fire-880 # zpool status pool: neptune_rpool state: ONLINE scrub: none requested config: NAME STATE READ WRITE CKSUM neptune_rpool ONLINE 0 0 0 mirror-0ONLINE 0 0 0 c1t0d0s0 ONLINE 0 0 0 c1t3d0s0 ONLINE 0 0 0 errors: No known data errors Now I want to add two more mirrors to that pool because the V880 has more drives to offer, that are not used at the moment. So I'd like to add in a mirror of c1t1d0 and c1t4d0 : # zpool add -f neptune_rpool c1t1d0 cannot label 'c1t1d0': EFI labeled devices are not supported on root pools. OKay .. I can live with that. # prtvtoc -h /dev/rdsk/c1t0d0s0 | fmthard -s - /dev/rdsk/c1t1d0s0 fmthard: New volume table of contents now in place. # prtvtoc -h /dev/rdsk/c1t0d0s0 | fmthard -s - /dev/rdsk/c1t4d0s0 fmthard: New volume table of contents now in place. # zpool add -f neptune_rpool c1t1d0s0 cannot add to 'neptune_rpool': root pool can not have multiple vdevs or separate logs So essentially there is no way to grow that zpool. Is this the case? -- Dennis ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] root pool can not have multiple vdevs ?
On Tue, Oct 27, 2009 at 4:39 PM, Dennis Clarke dcla...@blastwave.org wrote: So essentially there is no way to grow that zpool. Is this the case? There's the option of getting a bigger disk and do a send - receive. I'm guessing the restriction is necessary for simplicity sake to allow bootloaders work with zfs root. -- Fajar ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] root pool can not have multiple vdevs ?
On Tue, Oct 27, 2009 at 4:39 PM, Dennis Clarke dcla...@blastwave.org wrote: So essentially there is no way to grow that zpool. Is this the case? There's the option of getting a bigger disk and do a send - receive. I'm guessing the restriction is necessary for simplicity sake to allow bootloaders work with zfs root. You can grow your slice (I have done that) but that does require space after your slice. Casper ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Best practices for zfs create?
When I create a zfs filesystem there are lots of options. Which options are recommended? I use CIFS and then I choose mixedcasesensitivity=true. But it turns out that (due to a bug, fixed in b127) if I have non utf8 characters in the file name, I can not see the file in listings. So I should use utf8only which rejects files with non utf8 characters in it's file name. I didnt know of this option. Which other options do you use? -- 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] Best practices for zfs create?
Orvar Korvar wrote: When I create a zfs filesystem there are lots of options. Which options are recommended? That depends on what your needs are. The first consideration should be what kind of data you are storing. For a filesystem intended to be full of MP3 files it is wasteful of CPU and memory resources to enable compression. On the other hand if it is full of source code or ASCII text enabling compression could potentially improved performance - depending on the read and write access patterns. -- Darren J Moffat ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool getting in a stuck state?
Jeremy, I generally suspect device failures in this case and if possible, review the contents of /var/adm/messages and fmdump -eV to see if the pool hang could be attributed to failed or failing devices. Cindy On 10/26/09 17:28, Jeremy Kitchen wrote: Cindy Swearingen wrote: Hi Jeremy, Can you use the command below and send me the output, please? Thanks, Cindy # mdb -k ::stacks -m zfs ack! it *just* fully died. I've had our noc folks reset the machine and I will get this info to you as soon as it happens again (I'm fairly certain it will, if not on this specific machine, one of our other machines!) -Jeremy ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] resolve zfs properties default to actual value
Hi, how can I find out what the actual value when the default applies to a zfs property? # zfs get checksum mpool NAME PROPERTY VALUE SOURCE mpool checksum on default (In this particular case I know what the value is, either fletcher2 or fletcher4 depending on the build) But how can one find out in general what the actual value is? For any property. thank you, Frederik -- 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] resolve zfs properties default to actual value
Hi Frederik, In most cases, you can use the zfs get syntax below or you can use the zfs get all fs-name to review all current property settings. The checksum property is a bit different in that you need to review the zfs.1m man page checksum property description to determine the value of the default checksum property. You can use other methods to determine the current checksum value, but no easy way exists. The property descriptions and their defaults are up-to-date in table 6-1, here: http://docs.sun.com/app/docs/doc/817-2271/gazss?a=view Thanks, Cindy On 10/27/09 09:07, F. Wessels wrote: Hi, how can I find out what the actual value when the default applies to a zfs property? # zfs get checksum mpool NAME PROPERTY VALUE SOURCE mpool checksum on default (In this particular case I know what the value is, either fletcher2 or fletcher4 depending on the build) But how can one find out in general what the actual value is? For any property. thank you, Frederik ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] resolve zfs properties default to actual value
Thank you for the reply. I must admit that upon closer inspection alot of properties indeed do present the actual value. For the checksum property I used zdb - | grep fletcher to determine wether it was fletcher2 or fletcher4 was used for checksumming the filesystem. Using the OS build number to determine whether this was a fletcher2 filesystem isn't reliable. Perhaps this is a candidate for a RFE: change the value reported by zfs get checksum dataset from on to the actual used algorithm like fletcher2, fletcher4 or sha256 And what about compression? Off is off. But what about on? Is that gzip or lzjb? Same problem. Thanks, Frederik -- 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 code and fishworks fork
On Tue, Oct 27, 2009 at 2:35 AM, Bruno Sousa bso...@epinfante.com wrote: Hi all, I fully understand that within a cost effective point of view, developing the fishworks for a reduced set of hardware makes , alot, of sense. However, i think that Sun/Oracle would increase their user base if they make availabe a Fishwork framework certified only for a reduced set of hardware, ie : - it needs Western Digital HDD firmware version x.y.z - it needs a SAS/SATA controller from a specific brand, model and firmare ( LSI SAS1068E ) - if SSD's are used they need to be from vendor X with firmware Y - the system motherboard chipset needs to be from vendor X or Y and not from Z Within this possible landscape i'm pretty sure that alot more customers would pay for the Fishworks stack and support, given the fact that not all customers need aKa can afford, the Unified Storage platform from Sun. Anyway..Fishworks it's an awesome product! Congratulations for the extreme good job. Regards, Bruno You're making a very, very bad assumption that the price of Fishworks would be cheap for just the software. Sun hardware does not cost that much more than their competitors when it comes down to it. You should expect the software to make up the difference in price if they were to unbundle it. Heck, I would expect it to be MORE if they're forced into having to deal with third party vendors that are pointing fingers at software problems vs. hardware problems and wasting Sun support engineers valuable time. I think you'd find yourself unpleasantly surprised at the end price tag. --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Dumb idea?
I remember BFS (BeOS) did something very similar, it had extended metadata attributes akin to having a relational DB built-in. Very searchable, it also had tagging with callback notification on filechanges so you don't have to waste cycles with periodic checking. On 10/26/09 01:22, zfs-discuss-requ...@opensolaris.org wrote: -- Message: 2 Date: Sun, 25 Oct 2009 19:48:55 + From: j...@lentecs.com To: Orvar Korvar knatte_fnatte_tja...@yahoo.com,zfs Discuss zfs-discuss@opensolaris.org Subject: Re: [zfs-discuss] Dumb idea? Message-ID: 1672618364-1256500137-cardhu_decombobulator_blackberry.rim.net-13339056...@bda305.bisx.prod.on.blackberry Content-Type: text/plain; charset=Windows-1252 This actually sounds a little like what ms is trying to accomplish, in win7, with libraries. They will act as standard folders if you treat them as such. But they are really designed to group different pools of files into one easy place. You just have to configure it to pull from local and remote sources. I have heard it works well with win home server, and win7 networks. Its also similar to what google and the like are doing with their web crawlers. But I think this is something better left to run on top of the file system. Rather than integrated into the file system. A true database and crawling bot would seem to be the better method of implementing this. --Original Message-- From: Orvar Korvar Sender: zfs-discuss-boun...@opensolaris.org To: zfs Discuss Subject: [zfs-discuss] Dumb idea? Sent: Oct 24, 2009 8:12 AM Would this be possible to implement ontop ZFS? Maybe it is a dumb idea, I dont know. What do you think, and how to improve this? Assume all files are put in the zpool, helter skelter. And then you can create arbitrary different filters that shows you the files you want to see. As of now, you have files in one directory structure. This makes the organization of the files, hardcoded. You have /Movies/Action and that is it. But if you had all movies in one large zpool, and if you could programmatically define different structures that act as filters, you could have different directory structures. Programmatically defined directory structure1, that acts on the zpool: /Movies/Action Programmatically defined directory structure2: /Movies/Actors/AlPacino etc. Maybe this is what MS WinFS was about? Maybe tag the files? Maybe a relational database ontop ZFS? Maybe no directories at all? I dont know, just brain storming. Is this is a dumb idea? Or old idea? ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
Hi, I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. Why an increase in revenues? Well, i assume that alot of customers would buy the Fishworks to put into they XYZ high-end server. I often find alot of customers that say that it's far more easy to convince the Board of Directors to buy software rather than hardware or a appliance.. Maybe the first step could be the possibility to run Fishworks in any Sun server without the need to buy their Unified Storage server. In this way the support for the software and hardware would come from same vendor, therefore avoiding the dance between multiple vendor when it comes to fixing issues. Anyway, only time/market will say what's the best approach. Bruno Tim Cook wrote: On Tue, Oct 27, 2009 at 2:35 AM, Bruno Sousa bso...@epinfante.com mailto:bso...@epinfante.com wrote: Hi all, I fully understand that within a cost effective point of view, developing the fishworks for a reduced set of hardware makes , alot, of sense. However, i think that Sun/Oracle would increase their user base if they make availabe a Fishwork framework certified only for a reduced set of hardware, ie : * it needs Western Digital HDD firmware version x.y.z * it needs a SAS/SATA controller from a specific brand, model and firmare ( LSI SAS1068E ) * if SSD's are used they need to be from vendor X with firmware Y * the system motherboard chipset needs to be from vendor X or Y and not from Z Within this possible landscape i'm pretty sure that alot more customers would pay for the Fishworks stack and support, given the fact that not all customers need aKa can afford, the Unified Storage platform from Sun. Anyway..Fishworks it's an awesome product! Congratulations for the extreme good job. Regards, Bruno You're making a very, very bad assumption that the price of Fishworks would be cheap for just the software. Sun hardware does not cost that much more than their competitors when it comes down to it. You should expect the software to make up the difference in price if they were to unbundle it. Heck, I would expect it to be MORE if they're forced into having to deal with third party vendors that are pointing fingers at software problems vs. hardware problems and wasting Sun support engineers valuable time. I think you'd find yourself unpleasantly surprised at the end price tag. --Tim -- This message has been scanned for viruses and dangerous content by *MailScanner* http://www.mailscanner.info/, and is believed to be clean. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Sniping a bad inode in zfs?
I've have a single-fs, mirrored pool on my hands which recently went through a bout of corruption. I've managed to clean up a good bit of it but it appears that I'm left with some directories which have bad refcounts. For example, I have what should be an empty directory foo which, when you cd into it and ls -al, it shows a incorrect refcount for a empty directory: total 444 drwxr-xr-x 2 dalegusers 3 Aug 17 13:20 ./ drwx--x--x 64 dalegusers117 Aug 17 13:20 ../ Thus, attempts to remove this directory via rmdir fails with directory not empty and rm -rf gacks with File exists I can touch a new file in this dir and such, with the refcount incrementing to 4, and removing it poses no problem, either, with the refcount decrementing back to 3. However 3 is the wrong number. It should of course be only 2 (. and ..) Normally on UFS I would just take the 'nuke it from orbit' route and use clri to wipe the directory's inode. However, clri doesn't appear to be zfs aware (there's not even a zfs analog of clri in /usr/lib/fs/ zfs), and I don't immediately see an option in zdb which would help cure this. Any suggestions would be appreciated. /dale ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Sniping a bad inode in zfs?
On 27-Oct-09, at 1:43 PM, Dale Ghent wrote: I've have a single-fs, mirrored pool on my hands which recently went through a bout of corruption. I've managed to clean up a good bit of it How did this occur? Isn't a mirrored pool supposed to self heal? --Toby but it appears that I'm left with some directories which have bad refcounts. ... /dale ___ 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] Sniping a bad inode in zfs?
da...@elemental.org said: Normally on UFS I would just take the 'nuke it from orbit' route and use clri to wipe the directory's inode. However, clri doesn't appear to be zfs aware (there's not even a zfs analog of clri in /usr/lib/fs/ zfs), and I don't immediately see an option in zdb which would help cure this. Well, it might make things worse, but have you tried /usr/sbin/unlink ? I'm on Solaris-10, so don't know if that's still part of OpenSolaris. Regards, Marion ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
On Oct 27, 2009, at 12:35 AM, Bruno Sousa wrote: Hi all, I fully understand that within a cost effective point of view, developing the fishworks for a reduced set of hardware makes , alot, of sense. However, i think that Sun/Oracle would increase their user base if they make availabe a Fishwork framework certified only for a reduced set of hardware, ie : • it needs Western Digital HDD firmware version x.y.z • it needs a SAS/SATA controller from a specific brand, model and firmare ( LSI SAS1068E ) • if SSD's are used they need to be from vendor X with firmware Y • the system motherboard chipset needs to be from vendor X or Y and not from Z Do not underestimate the cost and complexity of maintaining compatibility matrices (I call them sparse matrices for a reason :-) Within this possible landscape i'm pretty sure that alot more customers would pay for the Fishworks stack and support, given the fact that not all customers need aKa can afford, the Unified Storage platform from Sun. There are competitors delivered as software-only: NexentaStor seems to be well designed and EON is progressing nicely. -- richard Anyway..Fishworks it's an awesome product! Congratulations for the extreme good job. Regards, Bruno Adam Leventhal wrote: With that said I'm concerned that there appears to be a fork between the opensource version of ZFS and ZFS that is part of the Sun/Oracle FishWorks 7nnn series appliances. I understand (implicitly) that Sun (/Oracle) as a commercial concern, is free to choose their own priorities in terms of how they use their own IP (Intellectual Property) - in this case, the source for the ZFS filesystem. Hey Al, I'm unaware of specific plans for management either at Sun or at Oracle, but from an engineering perspective suffice it to say that it is simpler and therefore more cost effective to develop for a single, unified code base, to amortize the cost of testing those modifications, and to leverage the enthusiastic ZFS community to assist with the development and testing of ZFS. Again, this isn't official policy, just the simple facts on the ground from engineering. I'm not sure what would lead you to believe that there is fork between the open source / OpenSolaris ZFS and what we have in Fishworks. Indeed, we've made efforts to make sure there is a single ZFS for the reason stated above. Any differences that exist are quickly migrated to ON as you can see from the consistent work of Eric Schrock. Adam -- Adam Leventhal, Fishworkshttp://blogs.sun.com/ahl ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. ___ 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] zpool getting in a stuck state?
Cindy Swearingen wrote: Jeremy, I generally suspect device failures in this case and if possible, review the contents of /var/adm/messages and fmdump -eV to see if the pool hang could be attributed to failed or failing devices. perusing /var/adm/messages, I see: Oct 22 05:06:11 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:11 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:11 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x1 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 lots of messages like that just prior to rsync warnings: Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:25:27 homiebackup10 rsyncd[776]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] I think the rsync warnings are indicative of the pool being hung. So it would seem that the bus is freaking out and then the pool dies and that's that? The strange thing is that this machine is way underloaded compared to another one we have (which has 5 shelves, so ~150TB of storage attached) which hasn't really had any problems like this. We had issues with that one when rebuilding drives, but it's been pretty stable since. looking at fmdump -eV, I see lots and lots of these: Oct 24 2009 05:02:54.098815545 ereport.io.scsi.cmd.disk.tran nvlist version: 0 class = ereport.io.scsi.cmd.disk.tran ena = 0x882108543f200401 detector = (embedded nvlist) nvlist version: 0 version = 0x0 scheme = dev device-path = /p...@0,0/pci8086,4...@5/pci1000,3...@0/s...@30,0 (end detector) driver-assessment = retry op-code = 0x28 cdb = 0x28 0x0 0x51 0x9c 0xa5 0x80 0x0 0x0 0x80 0x0 pkt-reason = 0x4 pkt-state = 0x0 pkt-stats = 0x10 __ttl = 0x1 __tod = 0x4ae2ecee 0x5e3ce39 always with the same device name. So, it would appear that the drive at that location is probably broken, and zfs just isn't detecting it properly? Also, I'm wondering if this is related to the thread just recently titled [zfs-discuss] SNV_125 MPT warning in logfile, as we're using the same controller that person mentions. We're going to order some beefier controllers with the next shipment, any suggestions on what to get? If we find that the new controllers work much better, we may even go as far as replacing the ones in the existing machines (or at least any machines experiencing these issues). We're not married to LSI, but we use LSI controllers in our webservers for the most part and they're pretty solid there (though admittedly those are hardware raid, rather than JBOD) Thanks so much for your help! -Jeremy signature.asc Description: OpenPGP digital signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. I'm afraid I don't see that argument at all; I think that the economics that you're advocating would be more than undermined by the necessarily higher costs of validating and supporting a broader range of hardware and firmware... - Bryan -- Bryan Cantrill, Sun Microsystems Fishworks. http://blogs.sun.com/bmc ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
On Tue, 27 Oct 2009, Bruno Sousa wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. Why an increase in revenues? Well, i assume that alot of customers would buy the Fishworks to put into they XYZ high-end server. Fishworks products (products that the Fishworks team developed) are designed, tweaked, and tuned for particular hardware configurations. It is not like general purpose OpenSolaris where the end user gets to experiment with hardware configurations and tunings to get the best performance (but might not achieve it). Fishworks engineers are even known to holler at the drives as part of the rigorous product testing. 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
Re: [zfs-discuss] zfs code and fishworks fork
Hi Adam, thank you for your precise statement. Be it only from an engineering standpoint, this is the kind of argumentation which I was expecting (and hoping for). I'm not sure what would lead you to believe that there is fork between the open source / OpenSolaris ZFS and what we have in Fishworks. I've caught myself thinking along these lines a couple of weeks ago before the fix for http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6604403 got integrated. I had thought they must have that fix in fishworks already, and I am glad to see that it's been put back into snv_125. At any rate, I think that the main selling point for 7xxxs is really the add on s/w and I believe that making the core technology openly available will strengthen the product rather than weakening it. Thank you all for your great work, Nils ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
On Oct 27, 2009, at 2:00 PM, Bryan Cantrill wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. I'm afraid I don't see that argument at all; I think that the economics that you're advocating would be more than undermined by the necessarily higher costs of validating and supporting a broader range of hardware and firmware... (Just playing Devil's Advocate here) There could be no economics at all. A basic warranty would be provided but running a standalone product is a wholly on your own proposition once one ventures outside a very small hardware support matrix. Perhaps Fishworks/AK would have a OpenSolaris edition - leave the bulk of the actual hardware support up to a support infrastructure that's already geared towards making wide ranges of hardware supportable - OpenSolaris/Solaris, after all, does allow that. Perhaps this could be a version of Fishworks that's not as integrated with what you get on a SUS platform; if some of the Fishworks functionality that depends on a precise hardware combo could be reduced or generalized, perhaps it's worth consideration. Knowing the little I do about what's going on under the hood of a SUS system, I wouldn't expect the version of Fishworks uses on the SUS systems to have 100% parity with a unbundled Fishworks edition - but the core features, by and large, would convey. /dale ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs send... too slow?
On Sun, October 25, 2009 03:45, Orvar Korvar wrote: It seems that zfs send... takes quite some time? 300GB takes 10 hours, this far. And I have in total 3TB to backup. This means it will take 100 hours. Is this normal? If I had 30TB to back up, it would take 1000 hours, which is more than a month. Can I speed this up? That seems pretty bad, I back up around 650GB to a USB-2 external drive (not the world's fastest!!!) in about 7 hours, last I checked the time. Is rsync faster? As I have understood it, zfs send.. gives me an exact replica, whereas rsync doesnt necessary do that, maybe the ACL are not replicated, etc. Is this correct about rsync vs zfs send? rsync doesn't seem to cover the ACLs, last I looked closely. Which is why I converted my working rsync-based version to a zfs send-receive version that only supports full backups and isn't automated yet. Grumble. (The ACLs are key for in-kernel CIFS, which is what drove me here.) -- David Dyer-Bennet, d...@dd-b.net; http://dd-b.net/ Snapshots: http://dd-b.net/dd-b/SnapshotAlbum/data/ Photos: http://dd-b.net/photography/gallery/ Dragaera: http://dragaera.info ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
On Oct 27, 2009, at 2:58 PM, Bryan Cantrill wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. I'm afraid I don't see that argument at all; I think that the economics that you're advocating would be more than undermined by the necessarily higher costs of validating and supporting a broader range of hardware and firmware... (Just playing Devil's Advocate here) There could be no economics at all. A basic warranty would be provided but running a standalone product is a wholly on your own proposition once one ventures outside a very small hardware support matrix. Perhaps Fishworks/AK would have a OpenSolaris edition - leave the bulk of the actual hardware support up to a support infrastructure that's already geared towards making wide ranges of hardware supportable - OpenSolaris/Solaris, after all, does allow that. Perhaps this could be a version of Fishworks that's not as integrated with what you get on a SUS platform; if some of the Fishworks functionality that depends on a precise hardware combo could be reduced or generalized, perhaps it's worth consideration. Knowing the little I do about what's going on under the hood of a SUS system, I wouldn't expect the version of Fishworks uses on the SUS systems to have 100% parity with a unbundled Fishworks edition - but the core features, by and large, would convey. Why would we do this? I'm all for zero-cost endeavors, but this isn't zero-cost -- and I'm having a hard time seeing the business case here, especially when we have so many paying customers for whom the business case for our time and energy is crystal clear... Hey, I was just offering food for thought from the technical end :) Of course the cost in man hours to attain a reasonable, unbundled version would have to be justifiable. If that aspect isn't currently justifiable, then that's as far as the conversation needs to go. However, times change and one day demand could very well justify the business costs. /dale ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool getting in a stuck state?
Jeremy Kitchen wrote: Cindy Swearingen wrote: Jeremy, I generally suspect device failures in this case and if possible, review the contents of /var/adm/messages and fmdump -eV to see if the pool hang could be attributed to failed or failing devices. perusing /var/adm/messages, I see: Oct 22 05:06:11 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:11 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:11 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x1 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 lots of messages like that just prior to rsync warnings: Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:25:27 homiebackup10 rsyncd[776]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] I think the rsync warnings are indicative of the pool being hung. So it would seem that the bus is freaking out and then the pool dies and that's that? The strange thing is that this machine is way underloaded compared to another one we have (which has 5 shelves, so ~150TB of storage attached) which hasn't really had any problems like this. We had issues with that one when rebuilding drives, but it's been pretty stable since. looking at fmdump -eV, I see lots and lots of these: Oct 24 2009 05:02:54.098815545 ereport.io.scsi.cmd.disk.tran nvlist version: 0 class = ereport.io.scsi.cmd.disk.tran ena = 0x882108543f200401 detector = (embedded nvlist) nvlist version: 0 version = 0x0 scheme = dev device-path = /p...@0,0/pci8086,4...@5/pci1000,3...@0/s...@30,0 (end detector) driver-assessment = retry op-code = 0x28 cdb = 0x28 0x0 0x51 0x9c 0xa5 0x80 0x0 0x0 0x80 0x0 pkt-reason = 0x4 pkt-state = 0x0 pkt-stats = 0x10 __ttl = 0x1 __tod = 0x4ae2ecee 0x5e3ce39 so doing some more reading here on the list and mucking about a bit more, I've come across this in the fmdump log: Oct 22 2009 05:03:56.687818542 ereport.fs.zfs.io nvlist version: 0 class = ereport.fs.zfs.io ena = 0x99eb889c6fe1 detector = (embedded nvlist) nvlist version: 0 version = 0x0 scheme = zfs pool = 0x90ed10dfd0191c3b vdev = 0xf41193d6d1deedc2 (end detector) pool = raid3155 pool_guid = 0x90ed10dfd0191c3b pool_context = 0 pool_failmode = wait vdev_guid = 0xf41193d6d1deedc2 vdev_type = disk vdev_path = /dev/dsk/c6t5d0s0 vdev_devid = id1,s...@n5000c50010a7666b/a parent_guid = 0xcbaa8ea60a3c133 parent_type = raidz zio_err = 5 zio_offset = 0xab2901da00 zio_size = 0x200 zio_objset = 0x4b zio_object = 0xa26ef4 zio_level = 0 zio_blkid = 0xf __ttl = 0x1 __tod = 0x4ae04a2c 0x28ff472e c6t5d0 is in the problem pool (raid3155) so I've gone ahead and offlined the drive and will be replacing it shortly. Hopefully that will take care of the problem! If this doesn't solve the problem, do you have any suggestions on what more I can look at to try to figure out what's wrong? Is there some sort of setting I can set which will prevent the zpool from hanging up the entire system in the event of a single drive failure like this? It's really annoying to not be able to log into the machine (and having to forcefully reboot the machine) when this happens. Thanks again for your help! -Jeremy signature.asc Description: OpenPGP digital signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. I'm afraid I don't see that argument at all; I think that the economics that you're advocating would be more than undermined by the necessarily higher costs of validating and supporting a broader range of hardware and firmware... (Just playing Devil's Advocate here) There could be no economics at all. A basic warranty would be provided but running a standalone product is a wholly on your own proposition once one ventures outside a very small hardware support matrix. Perhaps Fishworks/AK would have a OpenSolaris edition - leave the bulk of the actual hardware support up to a support infrastructure that's already geared towards making wide ranges of hardware supportable - OpenSolaris/Solaris, after all, does allow that. Perhaps this could be a version of Fishworks that's not as integrated with what you get on a SUS platform; if some of the Fishworks functionality that depends on a precise hardware combo could be reduced or generalized, perhaps it's worth consideration. Knowing the little I do about what's going on under the hood of a SUS system, I wouldn't expect the version of Fishworks uses on the SUS systems to have 100% parity with a unbundled Fishworks edition - but the core features, by and large, would convey. Why would we do this? I'm all for zero-cost endeavors, but this isn't zero-cost -- and I'm having a hard time seeing the business case here, especially when we have so many paying customers for whom the business case for our time and energy is crystal clear... - Bryan -- Bryan Cantrill, Sun Microsystems Fishworks. http://blogs.sun.com/bmc ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool getting in a stuck state?
Hi Jeremy, The ereport.io.scsi.cmd.disk.tran is describing connections problems to the /p...@0,0/pci8086,4...@5/pci1000,3...@0/s...@30,0 device. I think the .tran suffix is for transient. ZFS might be reporting problems with device as well, but if the zpool/zfs commands are hanging, then it might be difficult to get this confirmation. The zpool status command will report device problems. When a device in a pool fails, then I/O to the pool is blocked, reads might be successful. See the failmode property description in zpool.1m. Is this pool redundant? If so, you can attempt to offline this device until it is replaced. If you have another device available, you might replace the suspect drive and see if that solves the pool hang problem. Cindy On 10/27/09 12:04, Jeremy Kitchen wrote: Cindy Swearingen wrote: Jeremy, I generally suspect device failures in this case and if possible, review the contents of /var/adm/messages and fmdump -eV to see if the pool hang could be attributed to failed or failing devices. perusing /var/adm/messages, I see: Oct 22 05:06:11 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:11 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:11 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x1 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 lots of messages like that just prior to rsync warnings: Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:25:27 homiebackup10 rsyncd[776]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] I think the rsync warnings are indicative of the pool being hung. So it would seem that the bus is freaking out and then the pool dies and that's that? The strange thing is that this machine is way underloaded compared to another one we have (which has 5 shelves, so ~150TB of storage attached) which hasn't really had any problems like this. We had issues with that one when rebuilding drives, but it's been pretty stable since. looking at fmdump -eV, I see lots and lots of these: Oct 24 2009 05:02:54.098815545 ereport.io.scsi.cmd.disk.tran nvlist version: 0 class = ereport.io.scsi.cmd.disk.tran ena = 0x882108543f200401 detector = (embedded nvlist) nvlist version: 0 version = 0x0 scheme = dev device-path = /p...@0,0/pci8086,4...@5/pci1000,3...@0/s...@30,0 (end detector) driver-assessment = retry op-code = 0x28 cdb = 0x28 0x0 0x51 0x9c 0xa5 0x80 0x0 0x0 0x80 0x0 pkt-reason = 0x4 pkt-state = 0x0 pkt-stats = 0x10 __ttl = 0x1 __tod = 0x4ae2ecee 0x5e3ce39 always with the same device name. So, it would appear that the drive at that location is probably broken, and zfs just isn't detecting it properly? Also, I'm wondering if this is related to the thread just recently titled [zfs-discuss] SNV_125 MPT warning in logfile, as we're using the same controller that person mentions. We're going to order some beefier controllers with the next shipment, any suggestions on what to get? If we find that the new controllers work much better, we may even go as far as replacing the ones in the existing machines (or at least any machines experiencing these issues). We're not married to LSI, but we use LSI controllers in our webservers for the most part and they're pretty solid there (though admittedly those are hardware raid, rather than JBOD) Thanks so much for your help! -Jeremy ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
On Tue, Oct 27, 2009 at 2:13 PM, Dale Ghent da...@elemental.org wrote: On Oct 27, 2009, at 2:58 PM, Bryan Cantrill wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. I'm afraid I don't see that argument at all; I think that the economics that you're advocating would be more than undermined by the necessarily higher costs of validating and supporting a broader range of hardware and firmware... (Just playing Devil's Advocate here) There could be no economics at all. A basic warranty would be provided but running a standalone product is a wholly on your own proposition once one ventures outside a very small hardware support matrix. Perhaps Fishworks/AK would have a OpenSolaris edition - leave the bulk of the actual hardware support up to a support infrastructure that's already geared towards making wide ranges of hardware supportable - OpenSolaris/Solaris, after all, does allow that. Perhaps this could be a version of Fishworks that's not as integrated with what you get on a SUS platform; if some of the Fishworks functionality that depends on a precise hardware combo could be reduced or generalized, perhaps it's worth consideration. Knowing the little I do about what's going on under the hood of a SUS system, I wouldn't expect the version of Fishworks uses on the SUS systems to have 100% parity with a unbundled Fishworks edition - but the core features, by and large, would convey. Why would we do this? I'm all for zero-cost endeavors, but this isn't zero-cost -- and I'm having a hard time seeing the business case here, especially when we have so many paying customers for whom the business case for our time and energy is crystal clear... Hey, I was just offering food for thought from the technical end :) Of course the cost in man hours to attain a reasonable, unbundled version would have to be justifiable. If that aspect isn't currently justifiable, then that's as far as the conversation needs to go. However, times change and one day demand could very well justify the business costs. /dale The problem is, most of the things that make fishworks desirable are the things that wouldn't work. Want to light up a failed drive with an LED? Clustering? Timeouts for failed hardware? The fact of the matter is, people asking for this are people that aren't willing to spend the money that Sun would be asking for anyways. I mean, seriously, a 7110 is $10,000 LIST! Assuming you absolutely despise bartering on price, you can get the thing for 20% off just by using try and buy. If you're balking at that price, you wouldn't like the price of the software. No amount of but you don't have to support it is going to change that. I think you're failing to take into consideration the PR suicide it would be for Sun to offer fishworks on any platform people want, offer support contracts (that's the ONLY way this will make them money), and then turn around and tell people the reasony feature XYZ isnt' working is because their hardware isn't supported... oh, and they have no plans to ever add support either. I honestly can't believe this is even a discussion. What next, are you going to ask NetApp to support ONTAP on Dell systems, and EMC to support Enginuity on HP blades? Just because the underpinnings are based on an open source OS that supports many platforms doesn't mean this customized build can or ever should. And one last example... QLogic and Brocade FC switches run Linux... I wouldn't expect or ask them to make a version that I could run on a desktop full of HBA's to act as my very own FC switch even though it is entirely possible for them to do so. And just as a reminder... if you look back through the archives, I am FAR from a Sun fanboy... I just feel you guys aren't even grounded in reality when making these requests. --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool getting in a stuck state?
Jeremy, I can't comment on your hardware because I'm not familiar with it. If you have a storage pool with ZFS redundancy and one device fails or begins failing, then the pool keeps going, in a degraded mode but is generally available. You can try setting the failmode property to continue, which would allow reads to continue in case of a device failure, might prevent the pool from hanging. If offlining the disk or replacing the disk doesn't help, let us know. Cindy On 10/27/09 13:13, Jeremy Kitchen wrote: Jeremy Kitchen wrote: Cindy Swearingen wrote: Jeremy, I generally suspect device failures in this case and if possible, review the contents of /var/adm/messages and fmdump -eV to see if the pool hang could be attributed to failed or failing devices. perusing /var/adm/messages, I see: Oct 22 05:06:11 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:11 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:11 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x1 Oct 22 05:06:19 homiebackup10 scsi: [ID 365881 kern.info] /p...@0,0/pci8086,4...@1/pci1000,3...@0 (mpt1): Oct 22 05:06:19 homiebackup10 Log info 0x3108 received for target 5. Oct 22 05:06:19 homiebackup10 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 lots of messages like that just prior to rsync warnings: Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 05:55:29 homiebackup10 rsyncd[29746]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] Oct 22 06:10:29 homiebackup10 rsyncd[178]: [ID 702911 daemon.warning] rsync error: error in rsync protocol data stream (code 12) at io.c(453) [receiver=2.6.9] Oct 22 06:25:27 homiebackup10 rsyncd[776]: [ID 702911 daemon.warning] rsync: connection unexpectedly closed (0 bytes received so far) [receiver] I think the rsync warnings are indicative of the pool being hung. So it would seem that the bus is freaking out and then the pool dies and that's that? The strange thing is that this machine is way underloaded compared to another one we have (which has 5 shelves, so ~150TB of storage attached) which hasn't really had any problems like this. We had issues with that one when rebuilding drives, but it's been pretty stable since. looking at fmdump -eV, I see lots and lots of these: Oct 24 2009 05:02:54.098815545 ereport.io.scsi.cmd.disk.tran nvlist version: 0 class = ereport.io.scsi.cmd.disk.tran ena = 0x882108543f200401 detector = (embedded nvlist) nvlist version: 0 version = 0x0 scheme = dev device-path = /p...@0,0/pci8086,4...@5/pci1000,3...@0/s...@30,0 (end detector) driver-assessment = retry op-code = 0x28 cdb = 0x28 0x0 0x51 0x9c 0xa5 0x80 0x0 0x0 0x80 0x0 pkt-reason = 0x4 pkt-state = 0x0 pkt-stats = 0x10 __ttl = 0x1 __tod = 0x4ae2ecee 0x5e3ce39 so doing some more reading here on the list and mucking about a bit more, I've come across this in the fmdump log: Oct 22 2009 05:03:56.687818542 ereport.fs.zfs.io nvlist version: 0 class = ereport.fs.zfs.io ena = 0x99eb889c6fe1 detector = (embedded nvlist) nvlist version: 0 version = 0x0 scheme = zfs pool = 0x90ed10dfd0191c3b vdev = 0xf41193d6d1deedc2 (end detector) pool = raid3155 pool_guid = 0x90ed10dfd0191c3b pool_context = 0 pool_failmode = wait vdev_guid = 0xf41193d6d1deedc2 vdev_type = disk vdev_path = /dev/dsk/c6t5d0s0 vdev_devid = id1,s...@n5000c50010a7666b/a parent_guid = 0xcbaa8ea60a3c133 parent_type = raidz zio_err = 5 zio_offset = 0xab2901da00 zio_size = 0x200 zio_objset = 0x4b zio_object = 0xa26ef4 zio_level = 0 zio_blkid = 0xf __ttl = 0x1 __tod = 0x4ae04a2c 0x28ff472e c6t5d0 is in the problem pool (raid3155) so I've gone ahead and offlined the drive and will be replacing it shortly. Hopefully that will take care of the problem! If this doesn't solve the problem, do you have any suggestions on what more I can look at to try to figure out what's wrong? Is there some sort of setting I can set which will prevent the zpool from hanging up the entire system in the
Re: [zfs-discuss] zfs code and fishworks fork
are you going to ask NetApp to support ONTAP on Dell systems, well, ONTAP 5.0 is built on freebsd, so it wouldn't be too hard to boot on dell hardware. Hay, at least it can do aggregates larger than 16T now... http://www.netapp.com/us/library/technical-reports/tr-3786.html Rob ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] gigabyte iram
ma == Matthias Appel matthias.ap...@lanlabor.com writes: ma At the moment I'm considering using a Gigabyte iRAM as ZIL acard ans-9010 backs up to CF on power loss. also has a sneaky ``ECC emulation'' mode that rearranges non-ECC memory into blocks and then dedicates some bits to ECC, so you don't have to pay unfair prices for the 9th bit out of every eight. See the ``quick install'' pdf on acard's site which is written as a FAQ. it is a little goofy because it goes into this sort of ``crisis mode'' when you yank the power, where it starts using the battery and writing to the CF, which it doesn't do under normal operation. This exposes it to UPS-like silly problems if the battery is bad, or the CF is bad, or the CF isn't shoved in all the way, whatever, while something like stec that uses the dram and commits to flash continuously can report its failure before you have to count on it, which is particularly well-adapted to something like a slog that's not evenv used unless there's a power failure. but...the acard thing gives stec performance for intel prices! let us know how your .tw ramdisk scheme goes. ma (speed up sync. IO especially for NFS)? IIUC it will never give you more improvement than disabling the ZIL, so you should try disabling the ZIL and running tests to get an upper bound on the improvement you can expect. pgp4OsHgQ1WAE.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
Bruno Sousa wrote: Hi, I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. Why an increase in revenues? Well, i assume that alot of customers would buy the Fishworks to put into they XYZ high-end server. But in Bryan's blog.. http://blogs.sun.com/bmc/date/200811 "but one that also embedded an apt acronym: "FISH", Mike explained, stood for "fully-integrated software and hardware" -- which is exactly what we wanted to go build. I agreed that it captured us perfectly -- and Fishworks was born." Bruno I agree it would be great to have this sort of BUI on OpenSolaris, for example it makes CIFS integration in a AD/Windows shop a breeze, even I got it to work in a couple of minutes, but this would not be FISH. What the Fishworks team have shown is that Sun can make a admin GUI that is easy to use if they have a goal. Perhaps Oracle will help, but I see more lost sales of Solaris due it it being "difficult to manage" than any other reason. We may all not like MS Windows, but you can't say it's not easy to use. Compare it's RBAC implementation with Solaris. One is a straight forward tick GUI (admittedly not very extensible as far as I can see), the other a complete nightmare of files that need editing with vi! Guess which one is used the most? OpenSolaris is getting there, but 99% of all Sun's customers never see it as they are on Solaris 10. I recently bought a laptop just to run OpenSolaris and most things "just work"; it's my preferred desktop at home, but it still only does the simple stuff that Mac and Windows have done for years. Using any of the advance features however requires a degree in Systems Engineering. Ever wondered what makes Apple so successful? Apple makes FISH. www.eagle.co.nz This email is confidential and may be legally privileged. If received in error please destroy and immediately notify us. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Sun Flash Accelerator F20
jc == Jake Caferilla j...@tanooshka.com writes: jc But remember, this example system, has massive parallel jc scalability. jc I issue 2 read requests, both read requests return after 1 jc minute. yeah, it's interesting to consider overall software stacks that might have unnecessary serialization points. For example, Postfix will do one or two fsync()'s per mail it receives from the Interent into its internal queue, but I think it can be working on receiving many mails at once, all waiting on fsync's together. If the filesystem stack pushes these fsync's all the way down to the disk and serializes them, you'll only end up with a few parallel I/O transactions open, because almost everything postfix wants to do is ``write, sync, let me know when done.'' OTOH, if you ran three instances of Postfix on the same single box with three separate queue diectories on three separate filesystems/zpools, inside zones for example, and split the mail load among them using MX records or a load balancer, you might through this ridiculous arrangement increase the parallelism of I/O reaching the disk even on a system that serializes fsync's. but if ZFS is smart enough to block several threads on fsync at once, batch up their work to a single ZIL write-and-sync, then the three-instance scheme will have no benefit. anyway, though, I do agree it makes sense to quote ``latency'' as the write-and-sync latency, equivalent to a hard disk with the write cache off, and if you want to tell a story about how much further you can stretch performance through parallelism, quote ``sustained write io/s''. I don't think it's as difficult to quantify write performance as you make out, and obviuosly hard disk vendors are already specifying in this honest way otherwise their write latency numbers would be stupidly low, but it sounds like SSD vendors are sometimes waving their hands around ``oh it's all chips anyway we didn't know what you MEANT it's AMBIGUOUS is that a rabbit over there?'' giving dishonest latency numbers. Latency numbers for writes that will not survive power loss are *MEANINGLESS*. period. And that is worth complaining about. pgpyNHH4IuLbZ.pgp Description: PGP signature ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Sun Flash Accelerator F20
Miles Nordin wrote: For example, Postfix will do one or two fsync()'s per mail it receives from the Interent into its internal queue, but I think it can be working on receiving many mails at once, all waiting on fsync's together. If the filesystem stack pushes these fsync's all the way down to the disk and serializes them, you'll only end up with a few parallel I/O transactions open, because almost everything postfix wants to do is ``write, sync, let me know when done.'' OTOH, if you ran three instances of Postfix on the same single box with three separate queue diectories on three separate filesystems/zpools, inside zones for example, and split the mail load among them using MX records or a load balancer, you might through this ridiculous arrangement increase the parallelism of I/O reaching the disk even on a system that serializes fsync's. but if ZFS is smart enough to block several threads on fsync at once, batch up their work to a single ZIL write-and-sync, then the three-instance scheme will have no benefit. ZFS does exactly this. I demonstrate it on the SSD Discovery Days I run periodically in the UK. -- Andrew Gabriel ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] zpool import single user mode incompatible version
I know this is opensolaris and Solaris, but I'm stuck... I want to demonstrate to my client how to recover an unbootable system from a zfs snapshot. (Say some dope rm -rf /kernel/drv...) Running Solaris 10 U8 sparc. Normal procedures are boot cdrom -s (or boot net -s) zpool import rpool zfs rollback snapshot reboot and all is well I've done this before with earlier rev's of Sol 10. When I boot off Solaris 10 U8 I get the error that pool is formatted using an incompatible version. Status show the pool as Newer Version I know Update 8 has version 15, but it looks like the miniroot from the install media is only version 10. This is not good. Any advice? I am already thinking about installing U7 on my test box to demonstrate. Glad I haven't rolled out u8 into production. Thanks, Paul -- 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 code and fishworks fork
Trevor, Could not agree more, but not every costumer likes to have only a fancy GUI, even that this GUI is very well designed. However my point of view is based on the fact that the part of the software behind the Fishworks could be possible to install in other Sun servers, besides the 7xxx series. Regarding APPLE...well they have marketing gurus Bruno On Wed, 28 Oct 2009 09:47:31 +1300, Trevor Pretty wrote: Bruno Sousa wrote: Hi, I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. Why an increase in revenues? Well, i assume that alot of customers would buy the Fishworks to put into they XYZ high-end server. But in Bryan's blog.. http://blogs.sun.com/bmc/date/200811 [1] but one that also embedded an apt acronym: FISH, Mike explained, stood for fully-integrated software and hardware -- which is exactly what we wanted to go build. I agreed that it captured us perfectly -- and Fishworks was born. Bruno I agree it would be great to have this sort of BUI on OpenSolaris, for example it makes CIFS integration in a AD/Windows shop a breeze, even I got it to work in a couple of minutes, but this would not be FISH. What the Fishworks team have shown is that Sun can make a admin GUI that is easy to use if they have a goal. Perhaps Oracle will help, but I see more lost sales of Solaris due it it being difficult to manage than any other reason. We may all not like MS Windows, but you can't say it's not easy to use. Compare it's RBAC implementation with Solaris. One is a straight forward tick GUI (admittedly not very extensible as far as I can see), the other a complete nightmare of files that need editing with vi! Guess which one is used the most? OpenSolaris is getting there, but 99% of all Sun's customers never see it as they are on Solaris 10. I recently bought a laptop just to run OpenSolaris and most things just work; it's my preferred desktop at home, but it still only does the simple stuff that Mac and Windows have done for years. Using any of the advance features however requires a degree in Systems Engineering. Ever wondered what makes Apple so successful? Apple makes FISH. __ www.eagle.co.nz [2] This email is confidential and may be legally privileged. If received in error please destroy and immediately notify us. -- This message has been scanned for viruses and dangerous content by MAILSCANNER [3], and is believed to be clean. -- Bruno Sousa Links: -- [1] http://blogs.sun.com/bmc/date/200811 [2] http://www.eagle.co.nz/ [3] http://www.mailscanner.info/ -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. inline: smiley-laughing.gif___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool import single user mode incompatible version
On Tue, Oct 27, 2009 at 4:25 PM, Paul Lyons paulrly...@gmail.com wrote: I know this is opensolaris and Solaris, but I'm stuck... I want to demonstrate to my client how to recover an unbootable system from a zfs snapshot. (Say some dope rm -rf /kernel/drv...) Running Solaris 10 U8 sparc. Normal procedures are boot cdrom -s (or boot net -s) zpool import rpool zfs rollback snapshot reboot and all is well I've done this before with earlier rev's of Sol 10. When I boot off Solaris 10 U8 I get the error that pool is formatted using an incompatible version. Status show the pool as Newer Version I know Update 8 has version 15, but it looks like the miniroot from the install media is only version 10. This is not good. Any advice? I am already thinking about installing U7 on my test box to demonstrate. Glad I haven't rolled out u8 into production. Thanks, Paul Not to be a jerk, but is there a question in there? The system told you exactly what is wrong, and you seem to already know. You're booting from an old cd that has an old version of zfs. Grab a new iso. How would you expect a system that shipped with verison 10 of zfs to know what to do with version 15? --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
I just curious to see how much effort would it take to put the software of FISH running within a Sun X4275... Anyway..lets wait and see. Bruno On Tue, 27 Oct 2009 13:29:24 -0500 (CDT), Bob Friesenhahn bfrie...@simple.dallas.tx.us wrote: On Tue, 27 Oct 2009, Bruno Sousa wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. Why an increase in revenues? Well, i assume that alot of customers would buy the Fishworks to put into they XYZ high-end server. Fishworks products (products that the Fishworks team developed) are designed, tweaked, and tuned for particular hardware configurations. It is not like general purpose OpenSolaris where the end user gets to experiment with hardware configurations and tunings to get the best performance (but might not achieve it). Fishworks engineers are even known to holler at the drives as part of the rigorous product testing. Bob -- Bob Friesenhahn bfrie...@simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer,http://www.GraphicsMagick.org/ -- Bruno Sousa -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool import single user mode incompatible version
Tim Cook wrote: On Tue, Oct 27, 2009 at 4:25 PM, Paul Lyons paulrly...@gmail.com mailto:paulrly...@gmail.com wrote: When I boot off Solaris 10 U8 I get the error that pool is formatted using an incompatible version. You're booting from an old cd that has an old version of zfs. Grab a new iso. It might be that I can't read but does OP not state he is booting off Solaris 10 update 8 DVD? What can be newer than that one? If the miniroot really only supports ZFS v10 then this is indeed not good (unworkable/unusable/..) -- Dick Hoogendijk -- PGP/GnuPG key: 01D2433D + http://nagual.nl/ | SunOS 10u8 10/09 | OpenSolaris 2010.03 b125 + All that's really worth doing is what we do for others (Lewis Carrol) ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
Hi, Given the fact that i worked in the Healthcare industry and alot of my former customers wished to be able to run the former Sun NAS 5310 software in other hardware, i can see a interesting possible business case. In my former job, my customers liked the software used in the Sun StorageTek NAS appliance, but very few of them liked the concept of appliance..they prefer to have the same software in a non-appliance configuration, even if that means that SUN only has 1 server to support such a solution. Anyway i fully understand that the FISHworks is a combination of hw with software with some specific targets in mind, and for that i think FISHworks is the best of what the market has to offer these days... Bruno On Tue, 27 Oct 2009 18:58:19 +, Bryan Cantrill b...@eng.sun.com wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. I'm afraid I don't see that argument at all; I think that the economics that you're advocating would be more than undermined by the necessarily higher costs of validating and supporting a broader range of hardware and firmware... (Just playing Devil's Advocate here) There could be no economics at all. A basic warranty would be provided but running a standalone product is a wholly on your own proposition once one ventures outside a very small hardware support matrix. Perhaps Fishworks/AK would have a OpenSolaris edition - leave the bulk of the actual hardware support up to a support infrastructure that's already geared towards making wide ranges of hardware supportable - OpenSolaris/Solaris, after all, does allow that. Perhaps this could be a version of Fishworks that's not as integrated with what you get on a SUS platform; if some of the Fishworks functionality that depends on a precise hardware combo could be reduced or generalized, perhaps it's worth consideration. Knowing the little I do about what's going on under the hood of a SUS system, I wouldn't expect the version of Fishworks uses on the SUS systems to have 100% parity with a unbundled Fishworks edition - but the core features, by and large, would convey. Why would we do this? I'm all for zero-cost endeavors, but this isn't zero-cost -- and I'm having a hard time seeing the business case here, especially when we have so many paying customers for whom the business case for our time and energy is crystal clear... - Bryan -- Bryan Cantrill, Sun Microsystems Fishworks. http://blogs.sun.com/bmc ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Bruno Sousa -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool import single user mode incompatible version
On Tue, Oct 27, 2009 at 4:59 PM, dick hoogendijk d...@nagual.nl wrote: Tim Cook wrote: On Tue, Oct 27, 2009 at 4:25 PM, Paul Lyons paulrly...@gmail.commailto: paulrly...@gmail.com wrote: When I boot off Solaris 10 U8 I get the error that pool is formatted using an incompatible version. You're booting from an old cd that has an old version of zfs. Grab a new iso. It might be that I can't read but does OP not state he is booting off Solaris 10 update 8 DVD? What can be newer than that one? If the miniroot really only supports ZFS v10 then this is indeed not good (unworkable/unusable/..) Assuming he didn't accidentally burn the wrong media... and the 10u8 does in fact have 15 as the default version, but 10 as the mini-root (that sounds more than a bit odd to me), it's a matter of simply grabbing an opensolaris ISO instead to do the exact same thing. --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
Hi, Maybe during this emails you have missed the point that no one is requesting anything..we are just discussing a possible usage of FISHworks outside of the 7xxx series..more specific in other Sun Server. If i choose the personal point of view, my biggest wish is that i would love to run FISHwork in something else rather than a appliance..who knows, maybe Solaris 11 within a Sun X4275 instead of a 71110. Bruno On Tue, 27 Oct 2009 14:29:57 -0500, Tim Cook wrote: On Tue, Oct 27, 2009 at 2:13 PM, Dale Ghent wrote: On Oct 27, 2009, at 2:58 PM, Bryan Cantrill wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. I'm afraid I don't see that argument at all; I think that the economics that you're advocating would be more than undermined by the necessarily higher costs of validating and supporting a broader range of hardware and firmware... (Just playing Devil's Advocate here) There could be no economics at all. A basic warranty would be provided but running a standalone product is a wholly on your own proposition once one ventures outside a very small hardware support matrix. Perhaps Fishworks/AK would have a OpenSolaris edition - leave the bulk of the actual hardware support up to a support infrastructure that's already geared towards making wide ranges of hardware supportable - OpenSolaris/Solaris, after all, does allow that. Perhaps this could be a version of Fishworks that's not as integrated with what you get on a SUS platform; if some of the Fishworks functionality that depends on a precise hardware combo could be reduced or generalized, perhaps it's worth consideration. Knowing the little I do about what's going on under the hood of a SUS system, I wouldn't expect the version of Fishworks uses on the SUS systems to have 100% parity with a unbundled Fishworks edition - but the core features, by and large, would convey. Why would we do this? I'm all for zero-cost endeavors, but this isn't zero-cost -- and I'm having a hard time seeing the business case here, especially when we have so many paying customers for whom the business case for our time and energy is crystal clear... Hey, I was just offering food for thought from the technical end :) Of course the cost in man hours to attain a reasonable, unbundled version would have to be justifiable. If that aspect isn't currently justifiable, then that's as far as the conversation needs to go. However, times change and one day demand could very well justify the business costs. /dale The problem is, most of the things that make fishworks desirable are the things that wouldn't work. Want to light up a failed drive with an LED? Clustering? Timeouts for failed hardware? The fact of the matter is, people asking for this are people that aren't willing to spend the money that Sun would be asking for anyways. I mean, seriously, a 7110 is $10,000 LIST! Assuming you absolutely despise bartering on price, you can get the thing for 20% off just by using try and buy. If you're balking at that price, you wouldn't like the price of the software. No amount of but you don't have to support it is going to change that. I think you're failing to take into consideration the PR suicide it would be for Sun to offer fishworks on any platform people want, offer support contracts (that's the ONLY way this will make them money), and then turn around and tell people the reasony feature XYZ isnt' working is because their hardware isn't supported... oh, and they have no plans to ever add support either. I honestly can't believe this is even a discussion. What next, are you going to ask NetApp to support ONTAP on Dell systems, and EMC to support Enginuity on HP blades? Just because the underpinnings are based on an open source OS that supports many platforms doesn't mean this customized build can or ever should. And one last example... QLogic and Brocade FC switches run Linux... I wouldn't expect or ask them to make a version that I could run on a desktop full of HBA's to act as my very own FC switch even though it is entirely possible for them to do so. And just as a reminder... if you look back through the archives, I am FAR from a Sun fanboy... I just feel you guys aren't even grounded in reality when making these requests. --Tim -- This message has been scanned for viruses and dangerous content by MAILSCANNER [2], and is believed to be clean. -- Bruno Sousa Links: -- [1] mailto:da...@elemental.org [2] http://www.mailscanner.info/ -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] zpool failmode
Hi, I am trying to understand the behavior of zpool failmode=continue rpool. I've read the man page regarding to this and I understand that the default mode is set to wait. So If I set up my zfs pool to continue, in the case of loss of connectivity, what is this setting supposed to do? Does setting failmode=continue prevent the system not to panic when an event like loss of connectivity or pool failure occur? What does the EIO term refer to in the man page for this setting? Could somebody explain what really this setting does? Thanks Deniz -- 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] zpool failmode
On Tue, Oct 27, 2009 at 5:13 PM, deniz rende solarisw...@gmail.com wrote: Hi, I am trying to understand the behavior of zpool failmode=continue rpool. I've read the man page regarding to this and I understand that the default mode is set to wait. So If I set up my zfs pool to continue, in the case of loss of connectivity, what is this setting supposed to do? Does setting failmode=continue prevent the system not to panic when an event like loss of connectivity or pool failure occur? What does the EIO term refer to in the man page for this setting? Could somebody explain what really this setting does? Thanks Deniz wait causes all I/O to hang while the system attempts to retry the device. continue will cause the system to continue on as if nothing has changed. panic will cause the system to panic and core dump. The only real advantage I see in wait is that it will alert the admin to a failure rather quickly if you aren't checking the health of the system on a regular basis. --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
As far as I know, its an effort! Not just for x4275 specifically, but in general with any other x86 hardware and storage oriented software. A lot of work required to support a final solution as well. What Nexenta does with its version of NexentaStor is enabling third-party Partners to integrate software into a HW/SW solutions ready for production use. There is even a social network for Nexenta partners, where Partners talks to each other as well as to Nexenta experts and polishing their final NexentaStor solutions. Its a process and it works! List of Partners: http://www.nexenta.com/partners Bruno Sousa wrote: I just curious to see how much effort would it take to put the software of FISH running within a Sun X4275... Anyway..lets wait and see. Bruno On Tue, 27 Oct 2009 13:29:24 -0500 (CDT), Bob Friesenhahn bfrie...@simple.dallas.tx.us wrote: On Tue, 27 Oct 2009, Bruno Sousa wrote: I can agree that the software is the one that really has the added value, but to my opinion allowing a stack like Fishworks to run outside the Sun Unified Storage would lead to lower price per unit(Fishwork license) but maybe increase revenue. Why an increase in revenues? Well, i assume that alot of customers would buy the Fishworks to put into they XYZ high-end server. Fishworks products (products that the Fishworks team developed) are designed, tweaked, and tuned for particular hardware configurations. It is not like general purpose OpenSolaris where the end user gets to experiment with hardware configurations and tunings to get the best performance (but might not achieve it). Fishworks engineers are even known to holler at the drives as part of the rigorous product testing. 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] Moving an dataset zpool from a local zone to the global zone
I'm wondering if anyone has run into this before. I've got a zone that has a zpool added into it as a dataset. I need to remove the zpool from the zone and mount on the global zone directly. I can remove the dataset from the zone config and remove the zpool from the zone. But I am having trouble getting the zpool to mount again on the global zone. It imports and exports fine but will not let me change the mount point stating that it is part of a nonglobal zone. Any thoughts on this ? ie: server1 runs zone1 datapool is a zpool exported to zone1 as a dataset I am going to remove the zone and need server1 to host datapool and all it's filesystems. Also datapool is not the rootpool of the zone. 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] zfs code and fishworks fork
On Tue, Oct 27 at 18:58, Bryan Cantrill wrote: Why would we do this? I'm all for zero-cost endeavors, but this isn't zero-cost -- and I'm having a hard time seeing the business case here, especially when we have so many paying customers for whom the business case for our time and energy is crystal clear... - Bryan I don't have a need for a large 7110 box, my group's file serving needs are quite small. I decided on a Dell T610 running OpenSolaris, with half the drives populated now and half to be populated as we get close to filling them up. Pair of mirrored vdevs for performance, with an SSD cache. I'd have loved to have, instead, the nice fishworks gui interface to the whole thing, and if that existed on something like an X2270, that's what we would have bought instead of the Dell box. Ultimately, I wanted the simplicity of a Drobo, capable of saturating a Gig-E port or two, in an easy to maintain and administer system. One and a half out of three ain't bad, but Fishworks GUI on a 4-disk X2270 would have been a 3 for 3 solution I believe. We just can't afford to spend $8-10k to try a 7110 which is likely complete overkill for our needs, and we have no expectation of our business growing into it within the next two years. $2k was our absolute ceiling for a trial purchase, and I knew that if my OpenSolaris experiment didn't work out, I could just repurpose the Dell box with Debian, EXT3, software RAID and Samba and get a 75-80% solution. Yes, this may not make business sense for Sun-as-structured, but someone will figure out how to scratch that itch because it's real for a LOT of small businesses. They want that low cost entry into a business-grade NAS without having to build it themselves, something that's a step up from a whitebox 2-disk mirror from some no-name vendor who won't exist in 6 months. --eric PS: Not having enough engineers to support a growing and paying customer base is a *good* problem to have. The opposite is much, much worse. -- 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] zfs code and fishworks fork
On Wed, Oct 28, 2009 at 12:15 AM, Eric D. Mudama edmud...@bounceswoosh.orgwrote: On Tue, Oct 27 at 18:58, Bryan Cantrill wrote: Why would we do this? I'm all for zero-cost endeavors, but this isn't zero-cost -- and I'm having a hard time seeing the business case here, especially when we have so many paying customers for whom the business case for our time and energy is crystal clear... - Bryan I don't have a need for a large 7110 box, my group's file serving needs are quite small. I decided on a Dell T610 running OpenSolaris, with half the drives populated now and half to be populated as we get close to filling them up. Pair of mirrored vdevs for performance, with an SSD cache. I'd have loved to have, instead, the nice fishworks gui interface to the whole thing, and if that existed on something like an X2270, that's what we would have bought instead of the Dell box. Ultimately, I wanted the simplicity of a Drobo, capable of saturating a Gig-E port or two, in an easy to maintain and administer system. One and a half out of three ain't bad, but Fishworks GUI on a 4-disk X2270 would have been a 3 for 3 solution I believe. We just can't afford to spend $8-10k to try a 7110 which is likely complete overkill for our needs, and we have no expectation of our business growing into it within the next two years. $2k was our absolute ceiling for a trial purchase, and I knew that if my OpenSolaris experiment didn't work out, I could just repurpose the Dell box with Debian, EXT3, software RAID and Samba and get a 75-80% solution. Yes, this may not make business sense for Sun-as-structured, but someone will figure out how to scratch that itch because it's real for a LOT of small businesses. They want that low cost entry into a business-grade NAS without having to build it themselves, something that's a step up from a whitebox 2-disk mirror from some no-name vendor who won't exist in 6 months. --eric PS: Not having enough engineers to support a growing and paying customer base is a *good* problem to have. The opposite is much, much worse. So use Nexenta? --Tim ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs code and fishworks fork
Tim Cook wrote: PS: Not having enough engineers to support a growing and paying customer base is a *good* problem to have. The opposite is much, much worse. So use Nexenta? Got data you care about? Verify extensively before you jump to that ship.. :) ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss