Project

General

Profile

Bug #5792

Kernel panic while exporting ZFS

Added by Giorgos Mavrikas over 5 years ago. Updated over 5 years ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
-
Start date:
2015-04-03
Due date:
% Done:

0%

Estimated time:
Difficulty:
Medium
Tags:
needs-triage
Gerrit CR:

Description

I got this on a fresh OmniOS r151014.
Imported a usb disk containing a v28 ZFS pool.
zpool upgraded to feature flags.
pool export <pool> and I got a kernel panic with the trace at the end of this description.
I have saved the dump if anyone is interested in debugging it.


ffffff002051b910 vpanic()
ffffff002051b960 0xfffffffffba8b16d()
ffffff002051b990 range_tree_destroy+0x72(ffffff04f5522380)
ffffff002051b9d0 metaslab_fini+0xc6(ffffff04f54f6780)
ffffff002051ba10 vdev_metaslab_fini+0x56(ffffff04f4438300)
ffffff002051ba50 vdev_free+0x69(ffffff04f4438300)
ffffff002051ba90 vdev_free+0x4b(ffffff04f3000d40)
ffffff002051bad0 spa_unload+0xc6(ffffff04f4b9a000)
ffffff002051bb50 spa_export_common+0x115(ffffff04f4443000, 1, 0, 0, 0)
ffffff002051bb80 spa_export+0x2a(ffffff04f4443000, 0, 0, 0)
ffffff002051bbd0 zfs_ioc_pool_export+0x3e(ffffff04f4443000)
ffffff002051bc80 zfsdev_ioctl+0x4a7(10700000000, 5a03, 8042a80, 100003, ffffff04f26b0188, ffffff002051be68)
ffffff002051bcc0 cdev_ioctl+0x39(10700000000, 5a03, 8042a80, 100003, ffffff04f26b0188, ffffff002051be68)
ffffff002051bd10 spec_ioctl+0x60(ffffff04ec605880, 5a03, 8042a80, 100003, ffffff04f26b0188, ffffff002051be68, 0)
ffffff002051bda0 fop_ioctl+0x55(ffffff04ec605880, 5a03, 8042a80, 100003, ffffff04f26b0188, ffffff002051be68, 0)
ffffff002051bec0 ioctl+0x9b(3, 5a03, 8042


Files

illumos5792.png (283 KB) illumos5792.png console screenshot of panic Joshua M. Clulow, 2015-08-13 05:18 AM

Related issues

Related to illumos gate - Bug #6292: exporting a pool while an async destroy is running can leave entries in the deferred treeClosedMatthew Ahrens2015-10-07

Actions
#1

Updated by Joshua M. Clulow over 5 years ago

I just had the same panic, after doing the same thing. I had an old pool (v28) that I upgraded to feature flags, copied in a bunch of incremental snapshots (it is a backup disk) and exported.

#2

Updated by Alek Pinchuk about 5 years ago

  • Related to Bug #6292: exporting a pool while an async destroy is running can leave entries in the deferred tree added

Also available in: Atom PDF