Bug #8582
openKernel panic on `zpool import` with attached pool with corrupted metadata
0%
Description
I have a ZFS pool I have created and used on a Debian box with zfs-dkms. That pool got munged up during a resilver operation to a point where it can't be imported anymore on that system. In the hopes to recover some data from the pool I started trying to import the pool on other OSes using OpenZFS. I connected the pool to a VM running an updated instance of OpenIndiana Hipster. When running zpool import
, after a few seconds of disk activity the VM consistently reboots.
The pool consists of a single raidz2 vdev with 6 disks currently attached using external disk enclosures over USB passed into the VM. I'll keep the disks around unchanged in case I should do more tests with the pool or someone wants to look at the data on there.
Bear with me, this is the first time coming into contact with anything from the Solaris side of the universe. You may have to give me some hints at what parts of the system I need to look to provide you with necessary information.
I've attached the crash log extracted from the crash dump as described under How To Report Problems. I have the crash dump files on hand here if necessary.
Files
Updated by Bryan Horstmann-Allen almost 5 years ago
- Project changed from site to illumos gate