zpool -v status POOLNAME = coredump

42 views
Skip to first unread message

DaveGee

unread,
Oct 25, 2009, 8:09:51 PM10/25/09
to zfs-macos
Okay here's one that's been bugging me for some time now and I can't
seem to figure out what might be happening.

I have a setup that I've been having lots of fun playing with but have
found something odd that happens 'most' of the time.

Setup:

- I started my zfs adventure just slamming 2 Seagate 400GB SATA drives
together into a pool called MEDIA1
- No mirrors or nuthin just 2 400GB drives flying by the seat of my
pants as it were
- Got my hand on a Seagate 750 and slapped that sucker into the pool,
as advertised the MEDIA1 pool simply (and quickly) grew to a new
bigger size.

Neat 1.38TB's of storage... I have to manually empty the trash with:
sudo rm -r /Volumes/MEDIA1/.Trashes/501/* but other then that the
sucker is really nice to play with I've even thrown tons of junk files
to it just to see how full I can get it, I'm up to about 80% full.

One odd thing I've noticed...

I can do "zpool list" and get:

[CODE]
NAME SIZE USED AVAIL CAP HEALTH
ALTROOT
MEDIA1 1.41T 1.14T 275G 80% ONLINE -
[/CODE]

However MOST of the time when I do "zpool -v status MEDIA1" I get a
core-dump. I said MOSTLY only because I know that I have gotten a
successful report back some time in the past but I can't replicate the
success to any degree of regularity.

I've tried running it as 'su root command' and it didn't matter, I
tried becoming root via su and entering the root password and then
running the "zpool -v status MEDIA1" and I'd still usually get a core
dump however I am SURE I've gotten a positive response from that
command in the previous weeks/months but it's been a while I now.

So, in short, has anyone heard of "zpool -v status MEDIA1" causing a
core dump?

Info: OS X 10.5.8
ZFS: the binary version that was linked to on the MacZFS how to page.
109??!!? rings a bell I think.

Thanks for any input...

Dave

ca...@sepia.it

unread,
Oct 25, 2009, 9:10:03 PM10/25/09
to zfs-macos
Same problem here.

OS: 10.6.1
ZFS: 119 source

I damaged a pool on purpose with dd (only 1 byte in the middle) to see
what would happen on checksum errors. Then I read the files in the
pool and got an I/O error. After that I ran "zpool status" and got a
KP.

Maybe one of your drives is bad…

--

Interval Since Last Panic Report: -4 sec
Panics Since Last Report: 1
Anonymous UUID: C22117B0-
EB7D-0F52-90F0-5967B2FCC1EB

Sun Oct 25 22:14:59 2009
panic(cpu 0 caller 0x2a6ac2): Kernel trap at 0x004658c6, type 14=page
fault, registers:
CR0: 0x80010033, CR2: 0x0000000c, CR3: 0x5853c000, CR4: 0x000006e0
EAX: 0x00000000, EBX: 0x056102a0, ECX: 0x05c87b90, EDX: 0x00000000
CR2: 0x0000000c, EBP: 0x35c73d28, ESI: 0x00000200, EDI: 0x0b55c000
EFL: 0x00010206, EIP: 0x004658c6, CS: 0x00000008, DS: 0x72650010
Error code: 0x00000000

Backtrace (CPU 0), Frame : Return Address (4 potential args on stack)
0x35c73b48 : 0x21acfa (0x5ce650 0x35c73b7c 0x223156 0x0)
0x35c73b98 : 0x2a6ac2 (0x590a50 0x4658c6 0xe 0x590c1a)
0x35c73c78 : 0x29c968 (0x35c73c90 0x0 0x35c73d28 0x4658c6)
0x35c73c88 : 0x4658c6 (0xe 0x6750048 0x34f20010 0x2f200010)
0x35c73d28 : 0xf789ea (0x0 0x35c73dec 0x35c73dc8 0x3060ea)
0x35c73d48 : 0xf811e2 (0x0 0x0 0x35c73d68 0x0)
0x35c73d88 : 0x305301 (0x1f000000 0xcf205a20 0xb55c000 0x3)
0x35c73dc8 : 0x2f8779 (0x35c73de8 0x3 0x35c73e18 0x57d2ac)
0x35c73e18 : 0x2ed200 (0x5c87b90 0xcf205a20 0xb55c000 0x3)
0x35c73e78 : 0x46919d (0x4f0b280 0xcf205a20 0xb55c000 0x35c73f50)
0x35c73e98 : 0x494b01 (0x4f0b280 0xcf205a20 0xb55c000 0x35c73f50)
0x35c73f78 : 0x4ed85f (0x56102a0 0x56a5648 0x562d044 0x1)
0x35c73fc8 : 0x29d3fd (0x56a5644 0x0 0x10 0x6638f84)
Kernel Extensions in backtrace (with dependencies):
com.apple.filesystems.zfs(8.0)@0xf1d000->0xfdefff

BSD process name corresponding to current thread: zpool

Mac OS version:
10B504

Kernel version:
Darwin Kernel Version 10.0.0: Fri Jul 31 22:47:34 PDT 2009;
root:xnu-1456.1.25~1/RELEASE_I386
System model name: MacBookPro2,1 (Mac-F42189C8)

System uptime in nanoseconds: 19396965284593
unloaded kexts:
com.apple.driver.AppleFileSystemDriver 2.0 (addr 0xc0e000, size
0x12288) - last unloaded 123025973699
loaded kexts:
com.apple.filesystems.msdosfs 1.6 - last loaded 38462188308
com.apple.driver.AppleHIDKeyboard 1.1.5f1
com.apple.filesystems.zfs 8.0
com.apple.filesystems.autofs 2.1.0
com.apple.driver.AppleHWSensor 1.9.2d0
com.apple.driver.AppleUpstreamUserClient 3.0.5
com.apple.kext.ATIFramebuffer 6.0.2
com.apple.driver.SMCMotionSensor 3.0.0d4
com.apple.driver.AppleHDA 1.7.4a1
com.apple.ATIRadeonX1000 6.0.2
com.apple.driver.AirPort.Atheros 411.19.4
com.apple.driver.AppleBacklight 170.0.2
com.apple.kext.AppleSMCLMU 1.4.5d1
com.apple.Dont_Steal_Mac_OS_X 7.0.0
com.apple.iokit.CHUDUtils 201
com.apple.iokit.CHUDProf 212
com.apple.driver.AudioIPCDriver 1.1.0
com.apple.driver.AppleIRController 161
com.apple.driver.AppleIntelMeromProfile 19
com.apple.driver.ACPI_SMC_PlatformPlugin 3.4.0a20
com.apple.driver.AppleLPC 1.4.6
com.apple.driver.AppleUSBTrackpad 1.8.0b4
com.apple.driver.AppleUSBTCKeyEventDriver 1.8.0b4
com.apple.driver.AppleUSBTCKeyboard 1.8.0b4
com.apple.iokit.IOAHCIBlockStorage 1.5.0
com.apple.iokit.SCSITaskUserClient 2.5.1
com.apple.driver.AppleAHCIPort 2.0.0
com.apple.driver.AppleUSBHub 3.7.8
com.apple.BootCache 31
com.apple.AppleFSCompression.AppleFSCompressionTypeZlib 1.0.0d1
com.apple.driver.AppleSmartBatteryManager 160.0.0
com.apple.driver.AppleIntelPIIXATA 2.5.0
com.apple.driver.AppleUSBEHCI 3.7.5
com.apple.driver.AppleUSBUHCI 3.7.5
com.apple.iokit.AppleYukon2 3.1.14b1
com.apple.driver.AppleEFINVRAM 1.3.0
com.apple.driver.AppleRTC 1.3
com.apple.driver.AppleFWOHCI 4.3.4
com.apple.driver.AppleHPET 1.4
com.apple.driver.AppleACPIButtons 1.3
com.apple.driver.AppleSMBIOS 1.4
com.apple.driver.AppleACPIEC 1.3
com.apple.driver.AppleAPIC 1.4
com.apple.driver.AppleIntelCPUPowerManagementClient 90.0.0
com.apple.security.sandbox 0
com.apple.security.quarantine 0
com.apple.nke.applicationfirewall 2.0.11
com.apple.driver.AppleIntelCPUPowerManagement 90.0.0
com.apple.driver.AppleProfileReadCounterAction 17
com.apple.iokit.IOFireWireIP 2.0.3
com.apple.driver.AppleProfileTimestampAction 10
com.apple.driver.AppleProfileThreadInfoAction 14
com.apple.driver.AppleProfileRegisterStateAction 10
com.apple.driver.AppleProfileKEventAction 10
com.apple.driver.AppleProfileCallstackAction 20
com.apple.driver.DspFuncLib 1.7.4a1
com.apple.iokit.IO80211Family 300.20
com.apple.iokit.IONDRVSupport 2.0
com.apple.kext.ATI1600Controller 6.0.2
com.apple.kext.ATISupport 6.0.2
com.apple.iokit.IOSurface 73.0
com.apple.iokit.IOGraphicsFamily 2.0
com.apple.iokit.IOBluetoothSerialManager 2.2.1f7
com.apple.iokit.IOSerialFamily 10.0.2
com.apple.iokit.CHUDKernLib 203
com.apple.iokit.IOAudioFamily 1.7.0fc16
com.apple.kext.OSvKernDSPLib 1.3
com.apple.driver.AppleHDAController 1.7.4a1
com.apple.iokit.IOHDAFamily 1.7.4a1
com.apple.iokit.AppleProfileFamily 40
com.apple.driver.AppleSMC 3.0.1d2
com.apple.driver.IOPlatformPluginFamily 3.4.0a20
com.apple.driver.CSRUSBBluetoothHCIController 2.2.1f7
com.apple.driver.AppleUSBBluetoothHCIController 2.2.1f7
com.apple.iokit.IOBluetoothFamily 2.2.1f7
com.apple.driver.XsanFilter 402.1
com.apple.iokit.IOUSBHIDDriver 3.7.5
com.apple.driver.AppleUSBMergeNub 3.7.5
com.apple.driver.AppleUSBComposite 3.7.5
com.apple.iokit.IOSCSIMultimediaCommandsDevice 2.5.1
com.apple.iokit.IOBDStorageFamily 1.6
com.apple.iokit.IODVDStorageFamily 1.6
com.apple.iokit.IOCDStorageFamily 1.6
com.apple.iokit.IOATAPIProtocolTransport 2.5.0
com.apple.iokit.IOSCSIArchitectureModelFamily 2.5.1
com.apple.iokit.IOAHCIFamily 2.0.0
com.apple.iokit.IOUSBUserClient 3.7.5
com.apple.iokit.IOATAFamily 2.5.0
com.apple.iokit.IOUSBFamily 3.7.8
com.apple.iokit.IONetworkingFamily 1.8
com.apple.driver.AppleEFIRuntime 1.3.0
com.apple.iokit.IOFireWireFamily 4.1.7
com.apple.iokit.IOHIDFamily 1.6.0
com.apple.iokit.IOSMBusFamily 1.1
com.apple.kext.AppleMatch 1.0.0d1
com.apple.security.TMSafetyNet 6
com.apple.driver.DiskImages 281
com.apple.iokit.IOStorageFamily 1.6
com.apple.driver.AppleACPIPlatform 1.3
com.apple.iokit.IOPCIFamily 2.6
com.apple.iokit.IOACPIFamily 1.3.0
Model: MacBookPro2,1, BootROM MBP21.00A5.B08, 2 processors, Intel Core
2 Duo, 2.33 GHz, 2 GB, SMC 1.14f5
Graphics: ATI Radeon X1600, ATY,RadeonX1600, PCIe, 256 MB
Memory Module: global_name
AirPort: AirPort Extreme, Atheros 5416: 2.0.19.4
Bluetooth: Version 2.2.1f7, 2 service, 0 devices, 1 incoming serial
ports
Network Service: Ethernet, Ethernet, en0
Network Service: AirPort, AirPort, en1
Serial ATA Device: Hitachi HTS721010G9SA00, 93.16 GB
Parallel ATA Device: MATSHITADVD-R UJ-85J
USB Device: Built-in iSight, 0x05ac (Apple Inc.), 0x8501, 0xfd400000
USB Device: Keyboard Hub, 0x05ac (Apple Inc.), 0x1006, 0xfd800000
USB Device: USB Receiver, 0x046d (Logitech Inc.), 0xc51a, 0xfd830000
USB Device: Apple Keyboard, 0x05ac (Apple Inc.), 0x0220, 0xfd820000
USB Device: Bluetooth USB Host Controller, 0x05ac (Apple Inc.),
0x8205, 0x7d100000
USB Device: Apple Internal Keyboard / Trackpad, 0x05ac (Apple Inc.),
0x021a, 0x1d200000
USB Device: IR Receiver, 0x05ac (Apple Inc.), 0x8240, 0x5d200000

DaveGee

unread,
Oct 25, 2009, 10:43:32 PM10/25/09
to zfs-macos
Interesting... I'll have to check it out but I've been putting stuff
on the pool and playing it back... been just loading it up with video
from my EyeTV box so nothing to worry about if I loose it all I've
just ben recording 8pm to 11pm on 2 HD channels at a time... Fills
things up nice and easy.

I've played back the videos and the ones I've looked at all play fine
but perhaps its just certain files that will trigger it...

Odd

Alex Bowden

unread,
Oct 26, 2009, 8:22:11 AM10/26/09
to zfs-...@googlegroups.com

I've been streaming up to 4 simultaneous channels from 2 x dual
channel EyeTV boxes for since February and haven't seen a single error.

DaveGee

unread,
Oct 26, 2009, 3:07:43 PM10/26/09
to zfs-macos
Yep... and my pool is as far as I can tell by playing random files
from various dates that I hope might 'touch' all 3 drives (the 2x 400G
and 1x750G) and I can't find a single glitch in any of them... and yet
when I issue:

zpool -v status MEDIA1

A core dump will likely result. Oh and it happens RIGHT away... as in
no sooner do I press return and the core-dump gets displayed. So, it's
not like 'zpool status' it's 'working on stuff on the various drives'
and then finds something odd and dumps. The core dump literally
happens the same second I pressed return (give or take a millisecond
or ten) to execute the command. Its more of a curiosity more than
anything else.

I'd mention what hardware I was running but... I'd rather not self
incriminate,,, Sorry I've committed a mortal sin but I love to play
with hardware and the temptation was too great and besides I'm no
stranger to buying LOTS of Apple hardware... in fact I just ordered
TWO brand new 27" iMacs for my wife and myself and her iMac will get
handed down to her dad who was just starting to have trouble with the
old iMac PPC we bought for him one Christmas, and I could start
spouting off all the laptops and G4s etc etc etc... Yea Apple does
very well by me and I feel zero guilt in indulging my geek side with
this franken-mac.

ca...@sepia.it

unread,
Oct 26, 2009, 3:42:06 PM10/26/09
to zfs-macos
Dave, If you don't mind KPs, try "zpool clear MEDIA1" to clear error
counts and then "zpool status MEDIA1".

DaveGee

unread,
Oct 26, 2009, 6:34:29 PM10/26/09
to zfs-macos
Mind? Heck no! :)

Okay so:

zpool clear MEDIA1 caused zero issues ... I just was returned to the
command line without any indication at all.

zpool status MEDIA1 (core dump)

Is this odd or what?

Bill Winnett

unread,
Oct 26, 2009, 6:59:18 PM10/26/09
to zfs-...@googlegroups.com


I remember seeing this error happen to a pool of mine once. Turns
out that the zpool clear command did *not* clear the specific error
that was causing the status command to core dump on. I ended up
running a virtual machine loaded with a version of solaris on it, I
think the release was called milax, and I added the disks to the VM as
usb disks. The Solaris "zpool -v status" clearly showed the errors,
and "zpool clear" cleared them. Back in Leopard when I added the pool
"zpool -v status" was again a happy camper and did not core dump.

DaveGee

unread,
Oct 26, 2009, 7:57:50 PM10/26/09
to zfs-macos
Ah... never even crossed my mind to create a VM of Solaris... Hmmm I
have Parallels... gotta see what kind of damage I can do now! :) :)

Dave

Daniel

unread,
Oct 26, 2009, 11:19:00 PM10/26/09
to zfs-...@googlegroups.com
using virtualbox right now. works really well
--
"America was founded by men who understood that the threat of domestic tyranny is as great as any threat from abroad. If we want to be worthy of their legacy, we must resist the rush toward ever-increasing state control of our society. Otherwise, our own government will become a greater threat to our freedoms than any foreign terrorist."
- Ron Paul, Texas Straight Talk, May 31, 2004

DaveGee

unread,
Oct 27, 2009, 3:58:32 PM10/27/09
to zfs-macos


On Oct 26, 6:59 pm, Bill Winnett <bill.winn...@gmail.com> wrote:
>
> I think the release was called milax, and I added the disks to the VM as  
> usb disks.  The Solaris "zpool -v status" clearly showed the errors,  
> and "zpool clear" cleared them.  Back in Leopard when I added the pool  
> "zpool -v status" was again a happy camper and did not core dump.

Okay, I have virtualbox up and running (very cool btw) and have both
the latest Open-Solaris setup and I found a pre-configured disk image
with milax preinstalled. So now on to the next step... Is there any
way I can get the virtualbox to see the pools I created? They are all
SATA drives and I don't see of any way to get them to appear as USB
devices in the VBox config area.

Also how would one install/activate the web console for Open Solaris
and then my next questing would of course be how would I then add the
web based ZFS controls. I'm only 'slightly' comfortable with Solaris
but its not too different from the other UN*X I've worked on before.

Jason McNeil

unread,
Oct 27, 2009, 4:05:31 PM10/27/09
to zfs-...@googlegroups.com
You have to map it to a vmdk file using the VBoxManage command line tool that you then add to the virtual machine as it were just another virtual hard drive. I've had cases though where if the drive number (disk1 vs disk2) changes then the vmdk will be rejected. Once all the drives that are a part of your pool are available to the OpenSolaris setup it should be importable, either automatically or via a zpool import.

More detailed instructions can be found here, http://www.virtualbox.org/manual/UserManual.html#rawdisk

--jasonrm

426f6c74.55...@gmail.com

unread,
Oct 27, 2009, 10:04:40 PM10/27/09
to zfs-macos
Be very careful that you don't upgrade your zpool to an incompatible
version. You may not be able to do a zpool import back to os x.

Here's the zpool version on NexentaCore, an OpenSolaris with a Debian
package manager (apt-get is apt-clone on NexentaCore, which creates a
new boot environment from a snapshot; very, very cool):

<<

# uname -a
SunOS nc2myhost 5.11 NexentaOS_20090926 i86pc i386 i86pc Solaris

# zpool upgrade -v


This system is currently running ZFS pool version 18.

The following versions are supported:

VER DESCRIPTION
--- --------------------------------------------------------
1 Initial ZFS version
2 Ditto blocks (replicated metadata)
3 Hot spares and double parity RAID-Z
4 zpool history
5 Compression using the gzip algorithm
6 bootfs pool property
7 Separate intent log devices
8 Delegated administration
9 refquota and refreservation properties
10 Cache devices
11 Improved scrub performance
12 Snapshot properties
13 snapused property
14 passthrough-x aclinherit
15 user/group space accounting
16 stmf property support
17 Triple-parity RAID-Z
18 snapshot user holds
For more information on a particular version, including supported
releases, see:

http://www.opensolaris.org/os/community/zfs/version/N

Where 'N' is the version number.

>>

If you are using a version of Solaris with a later version of zfs than
os x (very, very likely), and you upgrade the pool from os x to work
on it in Solaris (also very likely), then you cannot export and import
the pool back into os x. Or, I should say, unless there is now a
command:

zpool downgrade ....

but I don't think there is.

I got burned with this when zfs first become possible on os x with the
developer distribution. At first, the zpool versions were the same,
but over time Solaris progressed fairly quickly, FreeBSD more slowly,
and os x... well, you know.

zfs has a similar command:

<<
# zfs upgrade -v

The following filesystem versions are supported:

VER DESCRIPTION
--- --------------------------------------------------------
1 Initial ZFS filesystem version
2 Enhanced directory entries
3 Case insensitive and File system unique identifier (FUID)
4 userquota, groupquota properties

For more information on a particular version, including supported
releases, see:

http://www.opensolaris.org/os/community/zfs/version/zpl/N

Where 'N' is the version number.
>>

Web gui's are great, and solaris with java has some great ones. But
you usually have to resort to the command line, so I suggest you
reacquaint yourself with man, apropos, less, grep, etc.:

<<
apropos zfs
apt-clone (8) - ZFS integrated APT package handling utility -
command-line interface
zdb (1m) - ZFS debugger
zfs (1m) - configures ZFS file systems
zpool (1m) - configures ZFS storage pools

zfs list | less

NAME USED AVAIL REFER MOUNTPOINT
syspool 5.23G 452G 28.5K legacy
syspool/rootfs-nmu-000 1.19G 452G 921M legacy
syspool/rootfs-nmu-001 57.5K 452G 828M legacy
syspool/rootfs-nmu-002 558M 452G 1.35G legacy
syspool/rootfs-nmu-003 177M 452G 1.41G legacy
syspool/rootfs-nmu-004 157M 452G 1.47G legacy
syspool/rootfs-nmu-005 234M 452G 1.59G legacy
syspool/rootfs-nmu-006 4.41M 452G 1.59G legacy
syspool/rootfs-nmu-007 8.16M 452G 1.62G legacy
syspool/rootfs-nmu-008 97.5M 452G 2.09G legacy
syspool/rootfs-nmu-009 1.26G 452G 2.10G legacy
syspool/rootfs-nmu-010 81.5K 452G 2.09G legacy
syspool/rootfs-nmu-011 211M 452G 2.11G legacy
syspool/rootfs-nmu-012 92.3M 452G 2.12G legacy
syspool/rootfs-nmu-013 294M 452G 2.30G legacy
syspool/rootfs-nmu-014 229M 452G 2.16G legacy
syspool/rootfs-nmu-015 778M 452G 2.36G legacy
zp14d 543G 142G 19K /zp14d
zp14d/arc1 543G 142G 22K /zp14d/arc1
zp14d/arc1/2005 135G 142G 105G /zp14d/arc1/2005
zp14d/arc1/2006 35.3G 142G 34.6G /zp14d/arc1/2006
zp14d/arc1/2007 373G 142G 323G /zp14d/arc1/2007
zp14t 842G 71.8G 28K /zp2cRM1TB
zp14t/InstallUpdate 29.4G 71.8G 29.4G /zp2cRM1TB/InstallUpdate
zp14t/Music 88.7G 71.8G 85.8G /zp2cRM1TB/Music
zp14t/Scans 7.13G 71.8G 7.13G /zp2cRM1TB/Scans
zp14t/ScansFromzpi1 274K 71.8G 2.55G /zp2cRM1TB/ScansFromzpi1
:
...

zfs list | grep zp14t

zp14t 842G 71.8G 28K /zp2cRM1TB
zp14t/InstallUpdate 29.4G 71.8G 29.4G /zp2cRM1TB/InstallUpdate
zp14t/Music 88.7G 71.8G 85.8G /zp2cRM1TB/Music
zp14t/Scans 7.13G 71.8G 7.13G /zp2cRM1TB/Scans
zp14t/ScansFromzpi1 274K 71.8G 2.55G /zp2cRM1TB/ScansFromzpi1
zp14t/archive 714G 71.8G 711G /zp2cRM1TB/archive
zp14t/zpi1BU 86K 71.8G 20K /zp2cRM1TB/zpi1BU
zp14t/zpiBU 2.55G 71.8G 2.55G /zp2cRM1TB/zpiBU

zpool status -v | less

pool: syspool
state: ONLINE
status: The pool is formatted using an older on-disk format. The pool
can
still be used, but some features are unavailable.
action: Upgrade the pool using 'zpool upgrade'. Once this is done,
the
pool will no longer be accessible on older software versions.
scrub: none requested
config:

NAME STATE READ WRITE CKSUM
syspool ONLINE 0 0 0
c1d0s0 ONLINE 0 0 0

errors: No known data errors

pool: zp14d
state: ONLINE
status: The pool is formatted using an older on-disk format. The pool
can
still be used, but some features are unavailable.
action: Upgrade the pool using 'zpool upgrade'. Once this is done,
the
pool will no longer be accessible on older software versions.
scrub: none requested
config:

NAME STATE READ WRITE CKSUM
zp14d ONLINE 0 0 0
mirror ONLINE 0 0 0
c2t0d0 ONLINE 0 0 0
:
...

zpool status -v | grep ONLINE

state: ONLINE
syspool ONLINE 0 0 0
c1d0s0 ONLINE 0 0 0
state: ONLINE
zp14d ONLINE 0 0 0
mirror ONLINE 0 0 0
c2t0d0 ONLINE 0 0 0
c2t1d0 ONLINE 0 0 0
state: ONLINE
zp14t ONLINE 0 0 0
mirror ONLINE 0 0 0
c5d0 ONLINE 0 0 0
c6d0 ONLINE 0 0 0

>>

This should work on just about any variant of solaris, from the
command line. You need to be root for most of this to work:

% su
password:
# <password for root>

or:

%sudo su
password: <your user password>

or:

%pfexec ... (avoid this unless you get to know it; try man pfexec;
cumbersome, but might be necessary;)

Does this help anyone? Do you want me to try to get zfs read write
running (again) on a PowerMac G5 on 10.5 server? It's fairly easy
with the older kext from the developer release, but I think you want
to roll your own kernel extension, with later bits for zfs?

Have you tried to modify and load the kernel modules for zfs from
FreeBSD? That might be a little closer to what you want, but I'm not
sure.

Let me know. It could take me a while, but I'll try to help when I
can. I had to get lots of zpools running quickly without so many bugs
and kernel panics, so I moved them off os x to solaris.

Good Luck,

Gordon


On Oct 27, 3:05 pm, Jason McNeil <ja...@jasonrm.net> wrote:
> You have to map it to a vmdk file using the VBoxManage command line tool
> that you then add to the virtual machine as it were just another virtual
> hard drive. I've had cases though where if the drive number (disk1 vs disk2)
> changes then the vmdk will be rejected. Once all the drives that are a part
> of your pool are available to the OpenSolaris setup it should be importable,
> either automatically or via a zpool import.
>
> More detailed instructions can be found here,http://www.virtualbox.org/manual/UserManual.html#rawdisk
>
> --jasonrm

DaveGee

unread,
Oct 27, 2009, 10:48:04 PM10/27/09
to zfs-macos
Okay so where to begin...

I didn't make it all the way I'm afraid...

Preface: In the very beginning I tried the commands referenced on the
page you linked and found I was getting errors (I think) because the
MEDIA1 ZFS pool was mounted in OS X.. so all the commands listed below
were issued when/while the MEDIA1 pool was not mounted.

=================================================
First I tried doing the command (I was not root nor using sudo):

VBoxManage internalcommands createrawvmd..... -rawdisk /dev/disk0 -
register

However I got a permissions error since I wasn't root...
=================================================


=================================================
Then I tried it again with the same:

sudo VBoxManage internalcommands createrawvmd..... -rawdisk /dev/disk0
-register

SUCCESS!: A vmdk file was created in ~/Library/VirtualBox/HardDisks
FAILURE!: However when I tried to add the drives in VirtualBox I got a
permission error
=================================================

=================================================
Then I set the owner of the fileX.vmdk to me:

sudo chown -R myname:myname file*.vmdk

SUCCESS!: I could add the drives to the virtual machine!!
FAILURE!: When I launched the VM I got a permission referencing the
file*.vmdk files.
=================================================

SO...

After a rebooting to clear the deck so to speak I tried this:

1 - zpool export -f MEDIA1

2a - sudo chmod a+rw /dev/disk0*
2b - sudo chmod a+rw /dev/disk2*
2c - sudo chmod a+rw /dev/disk3*

(yea I know this/above is not the thing to be doing but I had to try
something and a reboot put things back to the way they should be so no
harm no foul)

Then as me (no SU this time) I excuted the following:

For d0s2 (400GB)

VBoxManage internalcommands createrawvmdk -filename /Users/myname/
Library/VirtualBox/HardDisks/file0.vmdk -rawdisk /dev/disk0 -
partitions 1 -register

For d2 (750GB)

sudo VBoxManage internalcommands createrawvmdk -filename /Users/myname/
Library/VirtualBox/HardDisks/file2.vmdk -rawdisk /dev/disk2 -register

For d3s2 (400GB)

VBoxManage internalcommands createrawvmdk -filename /Users/myname/
Library/VirtualBox/HardDisks/file3.vmdk -rawdisk /dev/disk3 -
partitions 1 -register

Now if you notice d0 and d3 I'm specifying partition 1 but in
actuality on d0 and d3 it was s2 that was the main slice... but when I
issued some command that was indicated from the link above to
determine what partiton number I should be using it reported back 1 so
thats what I went with.

As for d2 ... well when I added that one I forgot to add the d0s2 so
instead I add the whole drive /dev/disk2 and ZFS didn't seem to care
it took the drive and added it without any complaints and while I
though I should have tried to fix my error and specify the drive the
correct way /dev/disk2s2 I didn't know enough about ZFS to do it so I
just left it be and started filling the drives and never gave it
another thought.

Okay so this time:

- The vmdk files got created without any issues!
- The vmdk files are now owned by ME!
- The vmdk files get accepted into the config file just fine (no
permission errors)
- The VM boots!! again no permission errors.
- My zpool doesn't show up...

I don't exactly know why but even as root I can't to an ls -al to my /
dev/ directory in open-solaris I don't know if its the way the os
behaves or if I downloaded a corrupt Virtual Box disk-image, or if its
something else...

Oh and while I see the disks are present when I over over the disk
icon in the frame of the VirtualBox window when I'm in the command
line they don't appear when I do a df -k and as stated above I can't
do a ls -al /dev/ since it just hangs... is their any other way to see
that solaris is actually seeing the drives... (read below because I
know its seeing some of them due to the fact that it KNEW the name of
my pool was MEDIA1.

I'm quite sure my 'fixes' are not the correct way to do things but it
was the only combonation of things that worked...

Oh interesting tidbit...

While in solaris I execute (as root):

% zpool export

And after quite a bit of time it eventually comes back with:

------------------------------------------------------------
pool: MEDIA1
id: a-long-id-number-was-here
state: UNAVAIL
status: One or more of the devices contains corrupted data
action: The pool cannot be imported du to damaged devices or data
see: http://www.sun.com/msg/ZFS-8000-5E
config:

MEDIA1 UNAVAIL insufficent replicas
c6t2d0p0 ONLINE
------------------------------------------------------------

Now when I quit the VM and reboot OS X and MEDIA1 mounts fine just
like it always has and I played a few files and all seemed well
enough... so clearly the pool wasn't damaged beyond repair so this too
makes me thing I did something wrong in the way I went about getting
the vmdk files to work...

I'm sure I'm doing something VERY wrong when it comes to "VBoxManage
internalcommands createrawvmdk...." but I just don't know what else I
could try.

Any pointers you can give me on how the command should be issued, as
root, doing a sudo (same thing really) , as me? and then how do I work
around the permission issues without resorting to sudo chmod a+rw /dev/
disk0*, sudo chmod a+rw /dev/disk2*, sudo chmod a+rw /dev/disk3* etc
etc etc...

Dave

On Oct 27, 4:05 pm, Jason McNeil <ja...@jasonrm.net> wrote:
> You have to map it to a vmdk file using the VBoxManage command line tool
> that you then add to the virtual machine as it were just another virtual
> hard drive. I've had cases though where if the drive number (disk1 vs disk2)
> changes then the vmdk will be rejected. Once all the drives that are a part
> of your pool are available to the OpenSolaris setup it should be importable,
> either automatically or via a zpool import.
>
> More detailed instructions can be found here,http://www.virtualbox.org/manual/UserManual.html#rawdisk

426f6c74.55...@gmail.com

unread,
Oct 27, 2009, 10:50:53 PM10/27/09
to zfs-macos
It's a real drag not having the old list imported directly here to
this new one. Maybe that's being done, I hope so.

On the old list, I posted the link to the opensolaris.org downloads
page:

http://hub.opensolaris.org/bin/view/Main/downloads

Of the various flavors listed there, I like Nexenta and Belenix the
best. But Nexenta is a commercial product, worth the money, but
expensive. There is, however, a free version at:

http://www.nexenta.org/

NCP 2 has a debian based package manager that works with zfs
snapshots, apt-clone. Belenix is a little more polished, easier to
get running with an XWindows gui, kde.

Gordon

DaveGee

unread,
Oct 27, 2009, 10:59:06 PM10/27/09
to zfs-macos


On Oct 27, 10:04 pm, "gwb0...@2Realms.com"
<426f6c74.55707269676...@gmail.com> wrote:
> Be very careful that you don't upgrade your zpool to an incompatible
> version.  You may not be able to do a zpool import back to os x.
>

You clearly had WAY more faith in me then you should have... lol

See my other post but long story short I can't seem to get the right
permutations to get my sata disks seen by VirtualBox...

One way was to run the commands with sudo but then when I tried to put
those files into the VM config I'd get permission errors... (which is
expected since I did the command via sudo) so I then tried to change
the ownership of the vmdk disk files back to me (from root) and then I
could add the vmdk disk files fine in the config but then when I tried
to boot the VM I'd then get a permission error almost as soon as the
system began to boot.

My workaround was to:

sudo chmod /dev/disk0* a+rw (same for /dev/disk2* /dev/disk3* )

Then use the commands indicated in the link above as me instead of
sudo and while the process got all the way thru booting without any
complaints about permissions the pool MEDIA1 wouldn't auto-mount, and
when I did a zpool import after a fair bit of time I got an error BUT
it did name my pool MEDIA1 (so something is being seen this time but
the error was such that the pool wasn't recoverable -- but when I
rebooted OS X bang MEDIA1 mounted without a hitch and I played a few
files and all was well...)...

I feel like I'm getting somewhere but I'm still not where I should
be...

BUT I thank you for your warnings and I will certainly heed them if/
when open-solaris actually finds my MEDIA1 pool and is able to mount
it...

Dave

426f6c74.55...@gmail.com

unread,
Oct 28, 2009, 2:05:47 AM10/28/09
to zfs-macos


On Oct 27, 9:59 pm, DaveGee <dgara...@gmail.com> wrote:
> On Oct 27, 10:04 pm, "gwb0...@2Realms.com"
>
> <426f6c74.55707269676...@gmail.com> wrote:
> > Be very careful that you don't upgrade your zpool to an incompatible
> > version.  You may not be able to do a zpool import back to os x.
>
> You clearly had WAY more faith in me then you should have... lol
>

You got farther than I ever did using VirtualBox with the "raw disk"
mode for SATA, so my faith is unshaken. I gave up on OpenSolaris on
VBox, and cobbled together a Gigabyte Intel Board, and a SATA/SAN Back
Plane to run Nexenta.

> See my other post but long story short I can't seem to get the right
> permutations to get my sata disks seen by VirtualBox...
>
> One way was to run the commands with sudo but then when I tried to put
> those files into the VM config I'd get permission errors... (which is
> expected since I did the command via sudo) so I then tried to change
> the ownership of the vmdk disk files back to me (from root) and then I
> could add the vmdk disk files fine in the config but then when I tried
> to boot the VM I'd then get a permission error almost as soon as the
> system began to boot.
>
> My workaround was to:
>
> sudo chmod /dev/disk0* a+rw (same for /dev/disk2* /dev/disk3* )
>
> Then use the commands indicated in the link above as me instead of
> sudo and while the process got all the way thru booting without any
> complaints about permissions the pool MEDIA1 wouldn't auto-mount, and
> when I did a zpool import after a fair bit of time I got an error BUT
> it did name my pool MEDIA1 (so something is being seen this time but
> the error was such that the pool wasn't recoverable -- but when I
> rebooted OS X bang MEDIA1 mounted without a hitch and I played a few
> files and all was well...)...
>
> I feel like I'm getting somewhere but I'm still not where I should
> be...
>

Have you tried mounting the disks on the host machine, and then using
iscsi? iscsi also allows a "raw disk" mode. Solaris has:

iscsitadm and iscsiadm

If you mount the disks on the host machine as targets, then you can
configure them as iscsi targets. Then the Virtual OS can use
iscsitadm and isciadm for the initiator (or target; I think iscsi
switches the meaning of target and initiator from the intuitive. Any
machine on your network can find the iscsi target, but the connection
speed limits how useful this can be.

zfs actually has an iscsi share command:

From:

http://docs.sun.com/app/docs/doc/819-5461/gebwq?a=view

<<
# zfs set shareiscsi=on tank/volumes/v2
# iscsitadm list target
Target: tank/volumes/v2
iSCSI Name: iqn.1986-03.com.sun:02:984fe301-c412-ccc1-cc80-
cf9a72aa062a
Connections: 0
>>

And, like for most things, Solaris has a few hundred densely written
pages on BigAdmin:

http://docs.sun.com/app/docs/doc/817-5093/fmvcd?a=view

It's pretty amazing that a file system (or something like it) would
have so many protocols built in.

> BUT I thank you for your warnings and I will certainly heed them if/
> when open-solaris actually finds my MEDIA1 pool and is able to mount
> it...
>
> Dave

You are welcome, and good luck. I'll try to get zfs read write
running again on the G5 Server, and try to replicate what you guys are
doing. It may take me a while.

Gordon
Reply all
Reply to author
Forward
0 new messages