NetBSD-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: zvol device weirdness after memory corruption?



On 11/11/25 17:19, Greg Troxel wrote:

Jeff Rizzo <riz%tastylime.net@localhost> writes:

# ccdconfig ccd0 1 /dev/zvol/dsk/tank/volumes/tempzvol
ccdconfig: ioctl (CCDIOCSET): /dev/rccd0d: Invalid argument
Are you sure your zpool has the same name, or did you do an
export/import stuff?
Not 100% sure I understand what you mean, but I didn't make a typo here - and the poolname is "tank" on both test systems.
Is the point of your usage of ccd, to get a "real disK" from a
partition/file?

Mostly, ccd is used here for a convenient example.  The problem showed up as "suddenly, my virtual disks don't attach in Xen"

Do you have a ccd config file?  I am guessing not.

Nope.  I only use it to be able to manipulate a zvol like a 'real' disk.  (thanks for the tip, mlelstv!)

Can you dd from the zvol?  gpt/fdisk?  fsck it?

I can dd to/from it (well... the raw device) and gpt works (raw or block). fsck doesn't work because I can't get at the partitions as wedges (which is the purpose ccd would serve).

I am somewhat surprised that ccd uses the block device but that's what
the man page says.


Yes, it works great on another system, and complains if the block device isn't used.


This is a head scratcher.  I did discover that wd3 is still showing recoverable errors (even though a zpool scrub completes successfully with no errors), so I'm probably going to replace that today, and also likely recreate the entire zpool because there's clearly something wrong.  I just thought it was weird, and wondered if anyone else might know what was up.

+j




Home | Main Index | Thread Index | Old Index