Hello!

On Sat, 08 Nov 2014 21:09:30 +0530 Ritesh Raj Sarraf <[email protected]> wrote:
On 10/20/2014 08:58 AM, dmitry wrote:
> Adding -x parameter to umountiscsi.sh
>
> And try to restart service:
> root@archive:~# /etc/init.d/open-iscsi restart
>
> Look on exerpt output(full report in attached file):
>
> Here (the cause of problem):
> + [ -d /sys/devices/platform/host4/iscsi_host ]
> + [ -d /sys/devices/platform/host4/session1/target4:0:0 ]
> + sed s/.*block\///
> + echo /sys/devices/platform/host4/session1/target*/*:*/block/*
> + BLOCK_DEV=*

I still think we do not need the patch. Because...

BLOCK_DEV=`echo "$BLOCK_FILE" | sed 's/.*block\///'`
if [ "${BLOCK_DEV}" = "*" ];then
log_warning_msg "iSCSI target without
block devices found"
continue
fi

Ideally you should have seen this warning message. But you are not
seeing this message. Which may be that you are using an older or
tampered version of the init script.


    I am sorry, but could you please be answered:

If code above present in umountiscsi.sh init script where can I found correct (not tampered) version? As far as I see from contents of package umountiscsi.sh script does not contain this code. I look at packages from any different repository (ftp.de.debian.org, ftp.fi.debian.org, ftp.ru.debian.org) and could not found it.



And the patch you provided is wrong. Because those block devices we are
looping for are not directories.

    Are you speaking about /dev/ contents? Am I wrong, that code:
"....
            for BLOCK_FILE in $SESSION_DIR/target*/*\:*/block/*; do
....
" -  loop for filesystem with directories?



> + awk /^\/dev\/*/ { print $2; }
> + DOS_PARTITIONS=/
> + umount /
> umount: /: device is busy.
> (In some cases useful info about processes that use
> the device is found by lsof(8) or fuser(1))
> + exit_status=1


--
Ritesh Raj Sarraf
RESEARCHUT - http://www.researchut.com
"Necessity is the mother of invention."


--
Dmitry Danilov.


--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]

Reply via email to