From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vincent Pelletier Date: Thu, 16 Aug 2018 12:11:16 +0000 Subject: Re: iscsi target: Let initiator decide whether it wants to authenticate target Message-Id: <20180816121116.130cc7ce@gmail.com> List-Id: References: <97479a0ccf17d8ad7a8ba7a0f7e8190da6ddc72e.1534334933.git.plr.vincent@gmail.com> In-Reply-To: <97479a0ccf17d8ad7a8ba7a0f7e8190da6ddc72e.1534334933.git.plr.vincent@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: target-devel@vger.kernel.org On Thu, 16 Aug 2018 00:00:35 -0500, Mike Christie wrote: > I just worry some users have set this and expect the extra layer of > checks. Now that you mention it, it is very possible indeed. The original code could help spot an initiator misconfiguration. > I can see how it is more convenient though. I think this is > something I really am not sure about because I have not worked on the > code for a long time. It is better if Nick were around. FWIW, I do not have a noteworthy use-case behind this - I merely have a home NAS with 2 LUNs exported. It is rather from a "I just noticed this going over the code and the RFC does not seem to require this" and a "I guess this confused/will confuse someone trying it out" point of view. > I saw some targets/initiators allow you to configure this type of thing > as optional where in that mode it works like in your patch. What about that? That would indeed be nicer than my patch. And then it can remain enforced by default. > I guess you can wait for other reviewers or maybe some distro packagers > to chime in too. I certainly can wait, yes. No worries. -- Vincent Pelletier