Re: [PATCH-v2 00/14] iscsi-target: iSCSI target v4.1.0-rc1 series initial merge

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sun, 2011-03-27 at 19:02 +0900, FUJITA Tomonori wrote:
> On Wed, 23 Mar 2011 23:59:40 -0700
> "Nicholas A. Bellinger" <nab@xxxxxxxxxxxxxxx> wrote:
> 
> > > > But reading the of the configfs layout is not the problem, yes..?
> > > 
> > > I don't think so. The user space doesn't cache the configfs info. So
> > > the rule is pretty simple.
> > 
> > It's more complicated than that.  We have to then start referencing
> > things like struct se_node_acl usage to userspace when locating an
> > explict NodeACL during login.  What happens if the explict NodeACL is
> > removed from configfs context while userspace code is running..?  What
> > happens if the entire TPG is removed during an active userspace login..?
> > It all has to be sychronized with a userspace daemon through a single
> > thread that needs to know about the network portals, TPG layout and
> > explict NodeACls in order to complete a iSCSI login.
> 
> Such can happen even if you do in kernel. During the login process,
> the configuration may change. You can't lock the configuration during
> the login process even if you do all in kernel. We need no
> synchronization at all with the user space approach.
> --

Not sure what you mean here by saying 'You can't lock the configuration
during the login process'.

The PATCH-v2 of iscsi-target code already provides the ability to
add/remove iSCSI network portals <-> iSCSI target portal group
assocations independently of other iSCSI TargetName+TargetPortalGroupTag
endpoints in a real-time M:N configuration via configfs regardless of
individual IPv6/IPv4 network portal or iSCSI login state.

Having to perform a level synchronization with an userspace iSCSI login
daemon for real-time configurable control plane of M:N iSCSI network
portals <-> iSCSI target portal group really does not make any sense
here.  This type of design change only adds extra unnecessary complexity
to existing code for in-kernel iscsi-target who's control plane is
already driven exclusively by userspace using mainline generic
target_core_fabric_configfs.c code.

--nab


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Index of Archives]     [SCSI Target Devel]     [Linux SCSI Target Infrastructure]     [Kernel Newbies]     [IDE]     [Security]     [Git]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux ATA RAID]     [Linux IIO]     [Samba]     [Device Mapper]
  Powered by Linux