Migrating to FreeNAS from Nexenta Core

Splendid · Dec 14, 2012

Hi guys,

I just imported my ZFS pool from my Nexenta Core system. All data sets are perfectly imported, but I have a problem with one of them. I can't open/view the contents of this volume. If I open via AFP the volume my FreeNas freezes. Same if I connect with FreeNas via SSH, I do a cd to this volume and freeze. My pools are healthy:

Code:

zpool status
  pool: data
 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.
  scan: scrub in progress since Sat Dec 15 00:25:58 2012
        18.0G scanned out of 2.11T at 32.1M/s, 19h0m to go
        0 repaired, 0.83% done
config:

	NAME                                            STATE     READ WRITE CKSUM
	data                                            ONLINE       0     0     0
	  mirror-0                                      ONLINE       0     0     0
	    gptid/cfa8a519-5f83-1ce1-a581-c9c8fccbd20a  ONLINE       0     0     0
	    gptid/f5d15303-c826-ad69-ee88-ad3a8fc2460c  ONLINE       0     0     0
	  mirror-1                                      ONLINE       0     0     0
	    gptid/13f52f4c-bb95-dfca-844b-db46d766c808  ONLINE       0     0     0
	    gptid/21ddf484-33ad-6f43-c191-9f7ce4070677  ONLINE       0     0     0

errors: No known data errors

As you can see above, it says I should upgrade my zpool, could that solve my problem? I'm scrubbing the pool now, maybe that helps. If there are any other options available I would like to try that first, before I upgrade my pool...

Thanks in advance!

Stephens · Dec 14, 2012

Since we don't know what's wrong, it's difficult to say what'll work. But I don't see why a pool upgrade would work when zpool status says the pool is fine. You'll know more when your scrube finishes. If you upgrade, you can't go back. If everything is fine with the pool, maybe there's something else corrupt with your FreeNAS installation (or computer).

Where'd you install FreeNAS? If to a USB flash drive, how big is it?

Can you mount/access the boot drive?

BobCochran · Dec 14, 2012

When you say

I can't open/view the contents of this volume.

can you explain how exactly you attempted to view the contents of the volume? Perhaps you should also post your dmesg output.

cyberjock · Dec 14, 2012

Have you tried a reboot? If not, don't do one now since it will restart your scrub.

You said you tried SSH and it sounded like if you go to /mnt/data and do an 'ls' you can't access the data and the system freezes...

Are you the same person that was in IRC last night? You cannot use the name 'data' for your zpool. I don't remember where I heard it to quote it, but if I remember correctly that's one of a small handful of names that cause conflicts. Can you go back to the Nexenta and rename the zpool then try FreeNAS again?

BobCochran · Dec 14, 2012

Well the zpool name 'data' appears to have worked on his Nexentra system. And it would seem to be a good name per the Oracle documentation.

From http://docs.oracle.com/cd/E19253-01/816-5166/zpool-1m/index.html :

zpool create [-fn] [-o property=value] ... [-O file-system-property=value] ... [-m mountpoint] [-R root] pool vdev ...

Creates a new storage pool containing the virtual devices specified on the command line. The pool name must begin with a letter, and can only contain alphanumeric characters as well as underscore (_), dash (-), and period (“.”). The pool names mirror, raidz, raidz1, raidz2, raidz3, spare, and log are reserved, as are names beginning with the pattern c[0-9]. The vdev specification is described in the “Virtual Devices” section.

...so I'm inclined to think we need a lot more detail from "Splendid".

cyberjock · Dec 14, 2012

'data' is a valid zpool name. But I thought there was a reason why FreeNAS can't handle it correctly.

Bob, you suck! LOL! Now I'm going to have to test it because I want to know for sure! :P

Edit: And here it is. I tried to create a zpool named 'data' and I got....

Dec 14 19:11:14 freenas kernel: GEOM_NOP: Device da1p2.nop created.
Dec 14 19:11:14 freenas manage.py: [middleware.exceptions:38] [MiddlewareError: Unable to create the pool: mountpoint '/data' exists and is not empty, use '-m' option to provide a different default, ]

So no. In short, you can't create a zpool named data.

To the OP: I think you'll need to take your zpool back to Nexenta and rename it to something else. Then I think it will work fine on FreeNAS. The fact that you were able to initiate the scrub tells me it did attempt to mount, is functional, and your data is safe. But the fact that another(hidden?) mountpoint exists tells me thats your problem.

BobCochran · Dec 14, 2012

You don't show your actual zpool create command. The error message is discussing your mount point, not your pool name. The error message is from manage.py, not from ZFS. I have my doubts about what you are asserting. If it worked fine on Nexentra, then it should work fine under any other ZFS implementation on any other system. There is so little real fact here that you can hardly tell him to go back and rename the pool. You don't have any information from the OP to base that on. What is needed first of all is real information from the OP about the nature of his or her problem.

Splendid · Dec 15, 2012

Thank you all for you replies.

I have multiple volumes on my NAS, I can access all of them, except one. This one is called 'Media'. When I try to access this one, via AFP or SSH, FreeNAS will freeze, shares unmount on my computer and my NAS isn't accessible via network. FreeNAS is running from a USB-stick (4GB).

When I try to connect to this volume (Media), I see the following message on my NAS:

Code:

panic: avl_find() succeeded inside avl_add()
cpuid = 0
KDB: enter: panic
[thread pid 10700 tid 100397 ]
Stopped at       kdb_enter+0x3b: movq     $0,0x74d242(%rip)

ProtoSD · Dec 15, 2012

data is reserved for FreeNAS and is the partition where it has the configuration database and reports etc.

You shouldn't need to go back to nexenta to rename your pool, just export it from the command line and reimport it with the new name (in the FAQ).

The media partition error is a whole different mystery. I'd try looking at "glabel status" or "gpart show". If it's not a ZFS partition that might be why, but not sure how that could happen.

Splendid · Dec 15, 2012

Just went back to Nexenta, imported the pool and everything works fine. Went back too FreeNAS, and stil no working Media volume. Rebooted to Nexenta, imported the pool with the name 'tank'. Rebooted again to FreeNAS imported pool with name 'tank'. Still no working 'Media' volume... :(

Response with "glabel status":

Code:

                                      Name  Status  Components
gptid/cfa8a519-5f83-1ce1-a581-c9c8fccbd20a     N/A  ada0p1
gptid/821686ee-e3d7-2ccf-883d-f63e079296eb     N/A  ada0p9
gptid/13f52f4c-bb95-dfca-844b-db46d766c808     N/A  ada1p1
gptid/7e63c15d-1a52-9bc5-9ee5-9deece330dcc     N/A  ada1p9
gptid/f5d15303-c826-ad69-ee88-ad3a8fc2460c     N/A  ada3p1
gptid/fe25d260-2f7e-6165-cc89-95488c799b66     N/A  ada3p9
gptid/21ddf484-33ad-6f43-c191-9f7ce4070677     N/A  ada4p1
gptid/733156db-82dc-3763-bc46-cf7caa663fc7     N/A  ada4p9
                             ufs/FreeNASs3     N/A  da1s3
                             ufs/FreeNASs4     N/A  da1s4
                            ufs/FreeNASs1a     N/A  da1s1a

Response with "gpart show":

Code:

=>        34  1953520061  ada0  GPT  (931G) [CORRUPT]
          34         222        - free -  (111k)
         256  1953503455     1  !6a898cc3-1dd2-11b2-99a6-080020736631  (931G)
  1953503711       16384     9  !6a945a3b-1dd2-11b2-99a6-080020736631  (8.0M)

=>        34  3906975677  ada1  GPT  (1.8T) [CORRUPT]
          34         222        - free -  (111k)
         256  3906959071     1  !6a898cc3-1dd2-11b2-99a6-080020736631  (1.8T)
  3906959327       16384     9  !6a945a3b-1dd2-11b2-99a6-080020736631  (8.0M)

=>        34  1953520061  ada3  GPT  (931G) [CORRUPT]
          34         222        - free -  (111k)
         256  1953503455     1  !6a898cc3-1dd2-11b2-99a6-080020736631  (931G)
  1953503711       16384     9  !6a945a3b-1dd2-11b2-99a6-080020736631  (8.0M)

=>        34  3906975677  ada4  GPT  (1.8T) [CORRUPT]
          34         222        - free -  (111k)
         256  3906959071     1  !6a898cc3-1dd2-11b2-99a6-080020736631  (1.8T)
  3906959327       16384     9  !6a945a3b-1dd2-11b2-99a6-080020736631  (8.0M)

=>     63  7829441  da0  MBR  (3.8G)
       63     4033       - free -  (2M)
     4096  7823360    1  !191  [active]  (3.7G)
  7827456     2048       - free -  (1.0M)

=>      63  15625153  da1  MBR  (7.5G)
        63   1930257    1  freebsd  [active]  (942M)
   1930320        63       - free -  (31k)
   1930383   1930257    2  freebsd  (942M)
   3860640      3024    3  freebsd  (1.5M)
   3863664     41328    4  freebsd  (20M)
   3904992  11720224       - free -  (5.6G)

=>      0  7823340  da0s1  EBR  (3.7G) [CORRUPT]
        0  7827456      1  !191  [active]  (3.7G)

=>      0  1930257  da1s1  BSD  (942M)
        0       16         - free -  (8.0k)
       16  1930241      1  !0  (942M)

Response with "gpart status":

Code:

  Name   Status  Components
ada0p1  CORRUPT  ada0
ada0p9  CORRUPT  ada0
ada1p1  CORRUPT  ada1
ada1p9  CORRUPT  ada1
ada3p1  CORRUPT  ada3
ada3p9  CORRUPT  ada3
ada4p1  CORRUPT  ada4
ada4p9  CORRUPT  ada4
 da0s1       OK  da0
 da1s1       OK  da1
 da1s2       OK  da1
 da1s3       OK  da1
 da1s4       OK  da1
 da0s5  CORRUPT  da0s1
da1s1a       OK  da1s1

I see a lot of corrupt components....any ideas?

cyberjock · Dec 15, 2012

protosd said:
data is reserved for FreeNAS and is the partition where it has the configuration database and reports etc.

Do you have anything to say for yourself Bob? Been around these parts before and seen problems with 'data' named zpools... plus the error. I think that's enough evidence to say that 'data' just doesn't work well with FreeNAS.. even without protosd's comment.

Splendid · Dec 15, 2012

Splendid said:

Ok, update on this. I could recover almost all components except one of course...

Code:

[root@SplendidNAS] ~# gpart status
  Name   Status  Components
ada0p1       OK  ada0
ada0p9       OK  ada0
ada1p1       OK  ada1
ada1p9       OK  ada1
ada3p1       OK  ada3
ada3p9       OK  ada3
ada4p1       OK  ada4
ada4p9       OK  ada4
 da0s1       OK  da0
 da1s1       OK  da1
 da1s2       OK  da1
 da1s3       OK  da1
 da1s4       OK  da1
 da0s5  CORRUPT  da0s1
da1s1a       OK  da1s1
[root@SplendidNAS] ~# gpart recover da0s1
gpart: pre-check failed: Operation canceled

The weird thing is, when I go to the 'Media' directory via SSH in the terminal, I can access subdirectories. So I can access 'Media/Music', when I do a 'ls -l' in this directory, it shows me the list of files and directories in 'Media/Music'. But when I do a 'ls -l' in 'Media' everything freezes. Aargh...

BobCochran · Dec 15, 2012

Thank you for the additional detail. I found this in the FreeNAS 8.3 Guide:

6.3.1 Auto Importing Volumes
...

NOTE: ...FreeNAS® 8.3.0 does not support compatibility with Nexenta pools or Linux fuse-zfs.

Also, we see that while a ZFS pool can be named 'data', FreeNAS apparently creates partitions which prevents the use of a zpool name of 'data'. There does not seem to be clear discussion of reserved names under FreeNAS in the guidebook. I found this quick mention:

Table 6.3a: Options When Creating a Volume
...
it is recommended to choose a name that will stick out in the logs (e.g. not data or freenas)

It might be a good idea to have a section in the Guide book on FreeNAS reserved names.

William Grzybowski · Dec 15, 2012

Splendid said:
Thank you all for you replies.

I have multiple volumes on my NAS, I can access all of them, except one. This one is called 'Media'. When I try to access this one, via AFP or SSH, FreeNAS will freeze, shares unmount on my computer and my NAS isn't accessible via network. FreeNAS is running from a USB-stick (4GB).

When I try to connect to this volume (Media), I see the following message on my NAS:

Code:
panic: avl_find() succeeded inside avl_add() cpuid = 0 KDB: enter: panic [thread pid 10700 tid 100397 ] Stopped at kdb_enter+0x3b: movq $0,0x74d242(%rip)

This is a known issue and should have been fixed in r12718.
Have you tried 8.3.0-RELEASE-p1, released yesterday?

Important Announcement for the TrueNAS Community.

Migrating to FreeNAS from Nexenta Core

Splendid

Cadet

Stephens

Patron

BobCochran

Contributor

cyberjock

Inactive Account

BobCochran

Contributor

cyberjock

Inactive Account

BobCochran

Contributor

Splendid

Cadet

ProtoSD

MVP

Splendid

Cadet

cyberjock

Inactive Account

Splendid

Cadet

BobCochran

Contributor

William Grzybowski

Wizard

Similar threads