Bug 321125 – sys-block/iscsitarget: ietd service won't stop if daemon was already killed

• Status: RESOLVED OBSOLETE
• Product: Gentoo Linux

[Description Jack Daniels 2010-05-22 21:48:27 UTC]

Start the ietd service; /etc/init.d/ietd start.

Kill the service; killall ietd.

Try and stop the service; /etc/init.d/ietd stop.
* Removing iSCSI Enterprise Target devices  [!!!]
* ietadm failed - Connection refused.

The and start the service again; /etc/init.d/ietd start
* WARNING:  ietd has already been started.

The issue is the init.d script. On line 86, it gets the output text of the ietadm app. If ietd was killed or crashed, ietadm will say "Connection refused." and return an error code. On line 88, we should hand-wave this past; it's a hack, but so is the whole stop() function:

Change
if [ $RETVAL -eq 0 ] && [ "$RETURN" != "something wrong" ] ; then
To
if [ $RETVAL -eq 0 ] && [ "$RETURN" != "something wrong" ] || [ "$RETURN" = 'Connection refused.' ] ; then

Reproducible: Always

Steps to Reproduce:

[Comment 1 Wormo (RETIRED) 2010-06-04 06:56:58 UTC]

Thanks for submitting your change to ietd init script -- assigning to maintainers

[Comment 2 Stefan Behte (RETIRED) 2010-06-19 17:08:05 UTC]

This is the wrong way to go. Upstream needs to fix ietd to send an other return code, if it fails.

[Comment 3 Richard Yao (RETIRED) 2013-01-03 11:18:04 UTC]

As of today, I have taken over primary maintainership of this package from base-system. I am reassigning this bug to myself.

[Comment 4 Jonas Stein 2018-10-04 21:23:55 UTC]

Any news? Is there a link to an upstream ticket?

[Comment 5 Michał Górny 2019-11-05 22:36:13 UTC]

removed.