On Mon, 2006-03-06 at 16:25 +0000, Stephen Howell wrote:
> Sure thing- didn't see that list.
sorry that i could not reproduce what u saw. i tried at my side with a
windows xp. it run ok after reboot normally and abnormally.
> I do a regular reboot most of the time, a couple of times I tried
> crashes but the result was the same.
> IET is a persistent target, it says "Reconnecting..." in the initiator
> for the PC and has the target listed after the reboot but the message
> never changes from "Reconnecting...", that is until I changed it to have
> 2 IQN entries with 1 LUN in each.
> So it is trying to reconnect, but hangs. If I then go into the
> initiator and tell it to log off the IET target and manually logon it is
> able to reconnect, although of course by then the cluster service has
> timed out and given up trying to acquire the disk resource.
could you run "tcpdump -s1600 -xlog" and redo this reboot. then send us
the log. also u ietd.conf file.
o, what is u ms ini version?
> This is running on MS Virtual Server by the way, I'm not aware of any
> networking issues that virtual server causes and I'm successfully
> running other *nix systems under it with network services (VS-FTP under
> debian and a BSD box running dummynet) and have had no trouble or odd
> -----Original Message-----
> From: Ming Zhang [mailto:mingz@...]
> Sent: 06 March 2006 15:42
> To: Stephen Howell
> Cc: iet-dev
> Subject: RE: [Iscsitarget-devel] Reconnection with multiple
> pls always cc to iet list.
> when u reboot it, u mean crash like reboot (simulate the failure by
> press power button directly) or normal reboot?
> have u add iet target as persistent target? if not, i think windows will
> not automatically connect iscsi at all.
> On Mon, 2006-03-06 at 15:18 +0000, Stephen Howell wrote:
> > I'm using 0.4.13, the problem occurs when running a single member of
> > the cluster on its own and rebooting it as well as when using multiple
> > connections.
> > Stephen
> > -----Original Message-----
> > From: Ming Zhang [mailto:mingz@...]
> > Sent: 06 March 2006 15:09
> > To: Stephen Howell
> > Cc: iet-dev
> > Subject: RE: [Iscsitarget-devel] Reconnection with multiple luns
> > periqnaddress
> > ic. if it is not in cluster, like a standalone windows, will u
> > reproduce this issue?
> > also which version of iet u use?
> > ming
> > On Mon, 2006-03-06 at 15:04 +0000, Stephen Howell wrote:
> > > After rebooting one of the cluster nodes (windows) to check cluster
> > > failover.
> > >
> > > -----Original Message-----
> > > From: Ming Zhang [mailto:mingz@...]
> > > Sent: 06 March 2006 14:55
> > > To: Stephen Howell
> > > Cc: iscsitarget-devel@...
> > > Subject: Re: [Iscsitarget-devel] Reconnection with multiple luns per
> > > iqnaddress
> > >
> > > On Fri, 2006-03-03 at 08:36 +0000, Stephen Howell wrote:
> > > > Hi there,
> > > >
> > > > I've installed you excellent project and it works very well. I'm
> > > > using it for a virtual cluster and it appears to handle multiple
> > > > clients just fine. However I have noted one small odd occurence.
> > > > When creating just one IQN and putting LUN 0 and LUN 1 under it
> > > > the Microsoft Initiator can connect ok, but after a reboot it
> > > > hangs saying
> > >
> > > reboot what? windows or iet?
> > >
> > > > "reconnecting" I can't see anything logged so I fiddled around
> > > > and found that using a seperate target IQN for each lun (one LUN
> > > > per target address) enabled clients to reconnect without any
> > > > Is
> > >
> > > > this intentional behaviour?
> > > >
> > > > cheers
> > > > Stephen Howell
> > >