search

fabioz23 / maczfs

Automatically exported from code.google.com/p/maczfs
Other
0 stars 0 forks source link

Zpool missing after reboot, part type now 0xEE Mavericks #130

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
First off had ZFS working find in Mountain Lion, stilled worked after upgrade 
to Mavericks.   
Redid ZFS for larger amount of drives (7).   When this problem was encounted 
ended up reformat drive, so this is a fresh install and still happening.

What steps will reproduce the problem?
1.  Installed the latest MacZFS-74.3.2b.pkg
2.  Followed the Getting Started guide, formatted drives check with diskutil 
list and all drive shown with zfs Type.
3. Setup up zpool no problems, zpool status no errors.
4.  Wrote data to drive.
5.  Rebooted.

What is the expected output? What do you see instead?
After reboot should see mounted zfs volume and my data.  
Instead, no mounted volume, get the mac message to reinitialize drive, I click 
ignore, get it several time for each drive. 

I ran diskutil list and now all the partitions are not zfs but 0xEE.

I can do a import, and drive is mounted but diskutil list still show 0xEE 

What version of the product are you using? On what operating system?
MacZFS-74.3.2b.pkg
Mac 10.9 Mavericks Now Clean install

Please provide any additional information below.

diskutil list
/dev/disk0
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *750.2 GB   disk0
   1:                       0xEE                         750.2 GB   disk0s1
/dev/disk1
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *1.0 TB     disk1
   1:                       0xEE                         1.0 TB     disk1s1
/dev/disk2
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *750.2 GB   disk2
   1:                       0xEE                         750.2 GB   disk2s1
/dev/disk3
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *1.0 TB     disk3
   1:                       0xEE                         1.0 TB     disk3s1
/dev/disk4
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:      GUID_partition_scheme                        *1.0 TB     disk4
   1:                        EFI EFI                     209.7 MB   disk4s1
   2:                  Apple_HFS Mac 2014                999.3 GB   disk4s2
   3:                 Apple_Boot Recovery HD             650.0 MB   disk4s3
/dev/disk5
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *1.0 TB     disk5
   1:                       0xEE                         1.0 TB     disk5s1
/dev/disk6
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *1.0 TB     disk6
   1:                       0xEE                         1.0 TB     disk6s1
/dev/disk7
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *1.0 TB     disk7
   1:                       0xEE                         1.0 TB     disk7s1

Original issue reported on code.google.com by mustn...@gmail.com on 11 Dec 2013 at 6:27

GoogleCodeExporter commented 9 years ago 
Thanks for the report.

The output from diskutil looks weird.  It lists all your disks as 
FDisk_partition, but these should all be GUID_partition_scheme.  While you can 
use FDisk partitions in a ZFS pool, auto mounting is only supported with the 
GUID partition scheme.

How did you performed step 3 in your list of steps?  Did you gave the partition 
names to the zpool command, or full disk names?

What is the output of "zpool status -v" after import?

Can you check if you have a /System/Library/Filesystems/zfs.fs entry?

Original comment by googlelogin@bjoern-kahl.de on 11 Dec 2013 at 10:39

GoogleCodeExporter commented 9 years ago 
Ok so I had step by step typed up for you and well it bombed.   In short....:)
Also of note, this is a early 2009 (4,1) mac pro dual xeon currently with 24 
gigs of ram, so I think I can rule out memory issues :)

Can you check if you have a /System/Library/Filesystems/zfs.fs entry?  YES

Prior to any reboot if I format the drive according to the Getting started 
guide:
diskutil partitiondisk /dev/disk2 GPTFormat ZFS %noformat% 100% 
Started partitioning on disk disk2 
Creating partition map 
[ + 0%..10%..20%..30%..40%..50%..60%..70%..80%..90%..100% ]  
Finished partitioning on disk disk2 
/dev/disk2 
   #:                   type name                size     identifier 
   0:  GUID_partition_scheme                    *9.4 GB   disk2 
   1:                    EFI                     200.0 MB disk2s1 
   2:                    ZFS                     9.0 GB   disk2s2 

I get the expected results no problem on all 7 drives.

I create the Zpool, I have done it with drives and partitions prior to 
reformatting Mavericks. 

After reformatting, I created the pools with just drive, thought I read that 
somewhere on the web.  
sudo zpool create data raidz2 disk0 disk1 disk2 disk3 disk5 disk6 disk7

After reboot of mac Pool is gone and DiskUtil list looks like my first post.
zpool import -f data - works and I see my raid and the data.

So my next step was to reformat drives again diskUtil looks correct with 
GUID_partition_scheme and zfs

sudo zpool create data raidz2 disk0s2 disk1s2 disk2s2 disk3s2 disk5s2 disk6s2 
disk7s2

My zpool was their no errors on status, my test data was intact.
Rebooted.

and mac rebooted, and rebooted, I could not see the crash message but just keep 
rebooting.

Next my brilliant idea was to go back to lion and start fresh, forgot to backup 
the doc I spent writing you this morning....

So got lion up and running installed MacZFS-74.3.2b.pkg, it like the install, 
diskutil looked great, did import no problem test data intact.

Reboot, crash. was able to take picture of screen, not sure if their is 
anything that would be helpful.

Reinstalled OS Lion....again :)  Installed MacZFS-74.3.0.pkg, diskutil list 
looked great.

Rebooted.
Zpool Mounted no import needed, data is their. SUCCESS 

I am not updated to Mountain Lion MacZFS-74.3.0.pkg is working fine and my test 
data is their.

It has got to be something in the new version.  

Granted my problem is solved by going back to Mountain Lion but I would like to 
be at Mavricks at some point.  
Is their anything I can provide you that may help anyone else?

James

Original comment by mustn...@gmail.com on 12 Dec 2013 at 8:56

GoogleCodeExporter commented 9 years ago 
just chiming in to say i'm running into the same issue on Mavericks. 
Tried reinstalling, reformatting drive, etc. 

All appears fine, read/writes are fast, but after restarting the machine the 
zpool disappears.
I've also tried installing the startup scripts.

Software  OS X 10.9 (13A603)
Processor  3.33 GHz 6-Core Intel Xeon
Memory  8 GB 1333 MHz DDR3 ECC

here is my zfs.fs file if that helps: http://cl.ly/T0aF

Original comment by m...@mike-tucker.com on 17 Dec 2013 at 7:23

Attachments:

GoogleCodeExporter commented 9 years ago 
Hello , 

I have OS X 10.8 and ZFS is working more than a year without any problem. After 
i install new 10.9 and install latest package MacZFS-74.3.2b.pkg and reboot i 
get Kernel Panic. 

To boot i login in safe mode and delete the zfs.kext from 
System/Library/Extensions . Tried older versions but they are not compatible 
with OS X 10.9.

After installation i can import my ZFS pool and everything is working normally 
until i reboot the machine ):

Original comment by valentin...@gmail.com on 17 Dec 2013 at 9:22

GoogleCodeExporter commented 9 years ago 
Dear valentin.hristev,

can you attach one of the panic logs you got after reboot?

Should be in /Library/Logs/DiagnosticReports/  or  /Library/Logs/CrashReporter/

The older version (since 74.3.0 at least, probably 74.2 also) are technically 
compatible with 10.9, only the installer was a safety check that stops the 
install process for 10.9.

Thanks

Björn

Original comment by googlelogin@bjoern-kahl.de on 17 Dec 2013 at 9:46

GoogleCodeExporter commented 9 years ago 
@ "mail@mike-tucker.com", comment #3:

Your zfs.fs is not the one for the latests MacZFS-74.3.2

Can you post the output of "pkgutil --pkgs | grep -e zfs" ?

Thanks

Björn

Original comment by googlelogin@bjoern-kahl.de on 17 Dec 2013 at 9:52

GoogleCodeExporter commented 9 years ago 
I'm also having a similar issue.  After a reboot, my ZFS pool doesn't show up 
and it's holding my photo library hostage.  I also upgraded to Mavericks and 
74.3.2, and all was fine until a recent reboot ( I don't reboot often, so I'm 
not sure when I fried things, but I'm guessing it was the 74.3.2 move).  My 
disks are below.  One of the two 4 TB drives I had in my pool shows up as ZFS 
and the other (drive 0 I'm guessing) looks odd.  

I'm using an Early 2009 Mac Pro with 2 2.93 quad Xeons and 12 GB of RAM on OS X 
10.9.1 (13B42).

davids-mac-pro:~ david$ diskutil list
/dev/disk0
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     FDisk_partition_scheme                        *4.0 TB     disk0
   1:                       0xEE                         2.2 TB     disk0s1
/dev/disk1
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:      GUID_partition_scheme                        *1.0 TB     disk1
   1:                        EFI EFI                     209.7 MB   disk1s1
   2:                  Apple_HFS Macintosh HD            837.9 GB   disk1s2
   3:                 Apple_Boot Recovery HD             650.0 MB   disk1s3
   4:       Microsoft Basic Data                         161.4 GB   disk1s4
/dev/disk2
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:      GUID_partition_scheme                        *4.0 TB     disk2
   1:                        EFI EFI                     209.7 MB   disk2s1
   2:          Apple_CoreStorage                         4.0 TB     disk2s2
   3:                 Apple_Boot Boot OS X               134.2 MB   disk2s3
/dev/disk3
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:                  Apple_HFS Internal Backup        *4.0 TB     disk3
/dev/disk5
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:     Apple_partition_scheme                        *2.0 TB     disk5
   1:        Apple_partition_map                         32.3 KB    disk5s1
   2:                  Apple_HFS My Book                 2.0 TB     disk5s3
/dev/disk6
   #:                       TYPE NAME                    SIZE       IDENTIFIER
   0:      GUID_partition_scheme                        *4.0 TB     disk6
   1:                        EFI EFI                     209.7 MB   disk6s1
   2:                        ZFS                         4.0 TB     disk6s2
davids-mac-pro:~ david$

Original comment by shinya...@gmail.com on 16 Feb 2014 at 2:08

GoogleCodeExporter commented 9 years ago 
The partition type showing as "0xEE" means the GPT structure on disk got 
damaged by something.  MacZFS is never writing to the GPT, so I don't know how 
you ended up in this situation.

Recovery will be tricky but doable.  I suggest we meet in IRC or on Skype for a 
step-by-step investigation and (hopefully) recovery.  Please contact me by mail 
for finding a time slot (I am in CET (UTC+1)).

Original comment by googlelogin@bjoern-kahl.de on 16 Feb 2014 at 9:44

GoogleCodeExporter commented 9 years ago 
Re: mustng69 (comment #2) and alentin.hristev (comment #4)
The reboot issue has been fixed recently, it was caused by a bug in MacZFS's 
ioctl handler, see issue 126.

Original comment by googlelogin@bjoern-kahl.de on 16 Feb 2014 at 9:51

GoogleCodeExporter commented 9 years ago 
Thanks for the response.  I am home today and tomorrow until early evening 
before I fly out for work and leave the computer in question. I am on EST 
(UTC-5). Let me know what time works for you.

Original comment by shinya...@gmail.com on 16 Feb 2014 at 4:36

GoogleCodeExporter commented 9 years ago 
OS: 10.9.4
Installed 74.3.3
DISK: Firewire 800 2TB drives /dev/disk4 /dev/disk6 used for simple zpool

All of the pools are gone and will not mount after reboot.
Looks like the same issue as before. Prior to this had 74.3.0 installed without 
any issue even after OS upgrade.

sh-3.2# cat Info.plist 
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" 
"http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
    <key>CFBundleDevelopmentRegion</key>
    <string>English</string>
    <key>CFBundleExecutable</key>
    <string>zfs</string>
    <key>CFBundleIconFile</key>
    <string></string>
    <key>CFBundleIdentifier</key>
    <string>org.maczfs.zfs.fs</string>
    <key>CFBundleInfoDictionaryVersion</key>
    <string>6.0</string>
    <key>CFBundleName</key>
    <string>zfs</string>
    <key>CFBundlePackageType</key>
    <string>KEXT</string>
    <key>CFBundleShortVersionString</key>
    <string>maczfs_74-3-3-0-ge37cf7b</string>
    <key>CFBundleSignature</key>
    <string>????</string>
    <key>CFBundleVersion</key>
    <string>74.3.3</string>
    <key>IOKitPersonalities</key>
    <dict/>
    <key>OSBundleLibraries</key>
    <dict>
        <key>com.apple.kpi.bsd</key>
        <string>8.0.0</string>
        <key>com.apple.kpi.libkern</key>
        <string>8.0.0</string>
        <key>com.apple.kpi.mach</key>
        <string>8.0.0</string>
        <key>com.apple.kpi.unsupported</key>
        <string>8.0.0</string>
    </dict>
</dict>
</plist>

Original comment by skylane...@gmail.com on 10 Oct 2014 at 6:20

GoogleCodeExporter commented 9 years ago 
Hi skylane67f, thanks for the report.

See below for instructions on how to get your pool back.

It is not entirely clear what the root cause of the problem is, but it seems 
something is reinitializing the disk with a MBR partitioning scheme (also 
called BIOS or DOS scheme) behind the users back.  It is not the MacZFS driver 
itself, since it does only know about GPT partition and does not contain code 
to write a proper MBR partition table.

*How to get your pool back:*

The basic idea is to recreate the correct GPT partition scheme.
This is possible, since the GPT partition scheme includes a backup copy, which 
is under normal conditions not damaged when some legacy software (re)formats a 
disk with the MBR scheme.

*Tools needed:*

* the gdisk utility, for example available from MacPorts or directly from here: 
http://sourceforge.net/projects/gptfdisk/ .  The right file is gdisk-0.8.8.pkg 
(or newer).
* the default Apple Terminal.app utility

You will need administrator access to the box, and I recommend doing the 
following at the physical console, not over a remote connection.

*The steps are:*

# open Terminal.app to get a command line.  Everything following will be done 
at the command line.

# determine the disk that needs fixing: run "diskutil list".  Note down the 
disk path, it will be something like "/dev/disk1". I will use "$disk" as a 
placeholder below.

# run sudo gdisk $disk  <-- replace with your disk path!

 This will probably complain about inconsistencies and that it loaded the backup copy, which is exactly what we want.

# with in gdisk, do: (All commands are single-letter commands and must be 
submitted with `<enter>`.  Most commands do not produce any output.)

 # type "p" (+ `<enter>`) to view the current (inconsistent) partition data.  Note: some gdisk versions show the loaded (healthy) backup copy here, not the corrupted primary copy, so the information is not reliable)

 # type "r" to go to the recovery menu

 # type "c" to copy the backup table over the primary table

 # type "v" to verify the table.  If the system complains about checksum errors, then abort.  The disk will then need a more thorough analysis and fixing.

 # usually "v" does not complain or just warns about some unallocated space.  That is perfectly fine.

 # type "w" to write the recovered partition table to disk.  This exits the gdisk tool.

# verify using "diskutil list" that the partition table has been fixed.

# Repeat with any other disk that needs fixing

# Done.  Reboot to make sure the Kernel will use the fixed table and not the 
corrupted old one.

Note:
Above steps are the shorted transcript from an IRC session to fix the disk from 
commenter 10.  Use with caution and do not write the new table to the disk, if 
you have doubts about the auto-recovered partition table as show by the "v" or 
"p" command.

Use at your own risk.

Original comment by googlelogin@bjoern-kahl.de on 11 Oct 2014 at 9:19