GJournal (hopefully) final patches.

Eric Anderson anderson at centtech.com
Thu Aug 17 15:02:47 UTC 2006

On 08/17/06 10:00, Pawel Jakub Dawidek wrote:
> On Thu, Aug 17, 2006 at 09:48:30AM -0500, Eric Anderson wrote:
>> On 08/17/06 09:08, Pawel Jakub Dawidek wrote:
>>> On Thu, Aug 17, 2006 at 08:08:56AM -0500, Eric Anderson wrote:
>>>> On 08/08/06 14:52, Pawel Jakub Dawidek wrote:
>>>>> Hi.
>>>>> I'm pleased to announce hopefully the final gjournal patches.
>>>> [..snip..]
>>>>> To create journaled UFS file system on ad0s1d partition:
>>>>> 	# gjournal load
>>>>> 	# gjournal label ad0s1d
>>>>> 	# newfs -J /dev/ad0s1d.journal
>>>>> 	# mount -o noatime,async /dev/ad0s1d.journal /mnt
>>>>> On reboot 'fsck_ffs -p /dev/ad0s1d.journal' will perform fast check. One
>>>>> can still run regular check by not giving the -p option.
>>>> I have a 10TB filesystem I created as above, and recently a crash caused it to now complain about the filesystem being dirty.  When I do an fsck -p, here's what I get:
>>>> # fsck -p /dev/label/vol10
>>> Do you have log from the boot before you saw this?
>>> I corrected one bug, but it was only possible to trigger when crash
>>> happened early in the boot process. Your crash was during or just after
>>> a boot maybe?
>> Yes, I have the logs.   What should I send you?
> I just need 'grep -i journal' from them.
>> Also, I just tried newfs'ing a new journaled device, and mounting it, then a few minutes later, crashing the system (not purposely though), and it too gives the same 
>> results as the other filesystem (same specs, 10Tb, etc).
>> The first crash was within about 30 minutes of booting, the second one was about the same time-frame.
> Maybe there was no writes?

There were no writes, absolutely.  The only part of the fs that was 
touched, was the superblock (marking clean/dirty).


Eric Anderson        Sr. Systems Administrator        Centaur Technology
Anything that works is better than anything that doesn't.

More information about the freebsd-fs mailing list