Re: [nslu2-linux] Mounting time for a hot plugged disk
- I think the beep is an unreliable cue, especially for people who don't
always work near the NSLU2. It certainly couldn't hurt, but I also don't
know how to make the unit beep when it's finished checking the disk
quotas. Maybe by wrapping the quota call with our own code or script?
If the only time the rw mounted drive becomes an issue is with unslinging,
then what about adding this to the unsling script?
if ! /bin/mount | /bin/grep -q "$targ type ext3 (rw)"; then
echo $1 is not mounted as read/write. Please try again later.
I forget if having the drive mounted rw is an issue when you use the
internal flash as the rootfs (since I never used 3.18), but since it's
been designated for advanced users for 4.20+, I'm not too worried about it
Can raise a SlugBug if we think this is a good fix.
--- Robert Hammond <rob.hammond@...> wrote:
> In message <355b7fc3050621144443090e95@...>, Kai Steinbach__________________________________________________
> <kai.steinbach@...> writes
> >thanks for picking this topic up again.
> >On 21/06/05, Robert Hammond <rob.hammond@...> wrote:
> >> So I though that I would start a discussion thread regarding the
> >> problems with hot plugged disks and mounting.
> >> It seems from various posts and with most recent firmware versions
> >> a number of new users (and some experienced users) have fallen fowl
> >> running console commands before a hot plugged disk is properly
> >> as rw (read/write).
> >I'm totally with you on this one. And I like your idea of beeping.
> Could it also be possible to echo some messages to the open console
> screen (if one is open, it probably would be) such as - hot plugged disk
> detected, mounting so please wait or mounting complete etc.
> >Did you actually see my earlier note on this topic?
> I have seen your original post but it seemed not to generate much of a
> >On 08/06/05, Kai Steinbach wrote:
> >> Developers,
> >> after Robert's great idea to provide a solution to the ipkg / fill
> >> flash topic, I can see another good opportunity here to improve the
> >> go-unslung process and make it more bulletproof for new users:
> >> On 08/06/05, chezdude2010 (heath) wrote:
> >> > If the disk isn't mounted RW you end up with tons of errors with
> >> > the unsling script when it tries to copy files to a RO partition.
> >> At the beginning of unsling, it would probably be possible to check
> >> the target partition is mounted (ro).
> >> If so, the script could also check if quotachecks are currently
> >> running. It could then advice the user to be patient and wait several
> >> minutes; IIRC mine takes more than 15 minutes - lots of small files.
> >> Or it could provide an option to - step 1 - kill the quota check and
> >> then - step 2 - remount the target partition in (rw).
> >> Let us know what you think. Makes sense? Any showstoppers that I
> >> missed? Feasible to implement?
> >What's easier to implement and more likely to catch all circumstances:
> >the beeping or the unsling changes?
> There does seem to be a check early in the current 5.5 unsling script
> that detects if the disk mount point exists but I think it will exist in
> both ro and rw modes.
> Robert Hammond
> [ Moderator Note: All new information should be recorded in the Wiki at
> http://www.nslu2-linux.org ]
> Yahoo! Groups Links
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around