#943347 open-iscsi: md-raid layers are not probed when deciding what sessions to stop during shutdown

Package:
open-iscsi
Source:
open-iscsi
Description:
iSCSI initiator tools
Submitter:
Christopher David Howie
Date:
2019-10-28 14:24:03 UTC
Severity:
normal
#943347#5
Date:
2019-10-23 17:34:32 UTC
From:
To:
I've created a VM to test the feasibility of a particular configuration
and wound up running into a case that open-iscsi does not consider
during shutdown: if an iSCSI session provides a device that is used as
part of an md-raid device, then the open-iscsi shutdown scripts will not
notice this and terminate the iSCSI session anyway.  This causes the
iSCSI device to be kicked from the raid device.  On boot, because the
device was kicked out, it will not be re-added automatically.  ("mdadm
$MD_DEVICE --re-add missing" fixes this, but must be run by hand.)

The current workaround I am using is to set
ISCSI_ROOT_KEEP_ALL_SESSIONS_AT_SHUTDOWN=1 in /etc/default/open-iscsi.

It would be nice if open-iscsi looked through md-raid layers and avoided
stopping sessions that provide a device used by _any_ active md-raid
device, as not doing so requires manually re-adding the iSCSI device to
the array next time it is assembled.
------------------------------------------------------------------------
                    IMPORTANT INFORMATION/DISCLAIMER

This document should be read only by those persons to whom it is
addressed.  If you have received this message it was obviously addressed
to you and therefore you can read it.

Additionally, by sending an email to ANY of my addresses or to ANY
mailing lists to which I am subscribed, whether intentionally or
accidentally, you are agreeing that I am "the intended recipient," and
that I may do whatever I wish with the contents of any message received
from you, unless a pre-existing agreement prohibits me from so doing.

This overrides any disclaimer or statement of confidentiality that may
be included on your message.

#943347#10
Date:
2019-10-28 14:18:57 UTC
From:
To:
Contorl: severity -1 wishlist

These are some of the use cases for which this functionality was added.

I have never done sw raid in all my years of interaction with
computers. At Big HW Vendor labs, it was mostly HW RAID.

The teardown of Linux (Stackable) Storage Layers is complicated and as
such, uncommon features like SW RAID got left out.

I'm willing to review and add this feature if someone can come up with
an implementation to integrate to what we have cooked so far.