Re: fsck in -current

On May 15, 2004, at 10:41 PM, Marc G. Fournier wrote:

> On Sat, 15 May 2004, Michael Hamburg wrote:
>
>> On May 15, 2004, at 9:08 PM, Marc G. Fournier wrote:
>>
>>>
>>> I'm seriously considering putting 5.x onto my next server, to take
>>> advantage of, if nothing else, the reduction in the GIANT LOCK 
>>> reliance
>>> ... one "concern" I have is how fsck works in 5.x ...
>>>
>>> Right now, on 4.x, I have an fsck running that has been going for 
>>> ~3hrs
>>> now:
>>>
>>> # date; ps aux | grep fsck
>>> Sat May 15 22:04:00 ADT 2004
>>> root    40 99.0  4.5 185756 185796  p0  R+    6:55PM 164:01.60 fsck 
>>> -y
>>> /dev/da0s1h
>>>
>>> and is in Phase 4 ...
>>>
>>> In 5.x, if I'm not mistaken, fsck's are backgrounded on reboot, so 
>>> that
>>> the system comes up faster ... but:
>>>
>>> a. wouldn't that slow down the fsck itself, since all the processes 
>>> on
>>> the
>>>    machine would be using CPU/memory?
>>
>> Yes.  You can probably renice it or something, though, and it wouldn't
>> take that much longer.
>>
>> Furthermore, if I recall correctly, it doesn't check as much after a
>> crash when soft-updates are enabled.  See below.
>>
>>> b. how long could fsck run in the background safely ... like, if I
>>>    rebooted a machine, fsck backgrounded and then all the processes
>>>    started up, is there a risk involved?
>>>
>>
>> No, fsck should be able to run in the background indefinitely with
>> basically no risk, so long as you have free space on your disk.  The
>> way that the latest fsck works is that it snapshots the drive, and 
>> then
>> checks the snapshot; the only type of error that it's expecting to 
>> find
>> is files which have been deleted but whose space has not been
>> reclaimed.  This space can be recovered without confusing running
>> processes.
>
> 'k ... fsck just started spewing messages to the screen, which is nice
> cause now I know its actually doing something:
>
> ZERO LENGTH DIR I=5159669  OWNER=root MODE=40755
> SIZE=0 MTIME=May 10 17:32 2004
> CLEAR? yes
>
> now, the ZERO LENGTH DIR is a result of using unionfs ... but would 
> that
> cause any issues with the snapshotting?  like, how much 'free space' 
> would
> you need, and what happens ifyou don't have enough? :(
>
I don't know about ZERO LENGTH DIRs... I've never had to deal with 
anything terribly interesting in an interactive fsck.

You don't need much space unless you have high turnover -- the default 
extra space in FFS that you can't allocate anyway should be plenty, but 
as far as I can tell, if you do run out of space, Very Bad Things 
Happen (tm).  Just what those Very Bad Things are probably depends on 
the release.

Mike