On Mon, 2008-11-03 at 15:50 -0500, Daniel Gruner wrote:
> On 11/3/08, Abhishek Kulkarni <[EMAIL PROTECTED]> wrote:
> >
> >  On Mon, 2008-11-03 at 15:16 -0500, Daniel Gruner wrote:
> >  > On 11/3/08, Abhishek Kulkarni <[EMAIL PROTECTED]> wrote:
> >  > >
> >  > >  On Mon, 2008-11-03 at 14:34 -0500, Daniel Gruner wrote:
> >  > >  > The group and passwd file are copied from the master (well, only 
> > some
> >  > >  > entries), but in any case they match.
> >  > >  >
> >  > >  > I was fearing you'd tell me about the public keys... :-(
> >  > >  >
> >  > >  > The answer to that is that I don't mount the home directories on the
> >  > >  > nodes by default, and I wasn't planning to do so, period.
> >  > >
> >  > >
> >  > > xcpufs skips over the user if it cannot find the user's public key in
> >  > >  the home directory mounted locally.
> >  > >
> >  >
> >  > Yeah, I noticed that.
> >  >
> >  > >
> >  > >  > In fact, in
> >  > >  > some sense having the password files and keys available on the nodes
> >  > >  > kind of defeats the security safeguards built into xcpu, especially 
> > if
> >  > >  > one uses the xcpufs -u option.  It would be far better to do as we
> >  > >  > discussed before:  let the scheduler assign the permissions.
> >  > >  >
> >  > >  > In the meantime (until the bjs port is available), this may still be
> >  > >  > the only way to do it...
> >  > >
> >  > >
> >  > > if you can pull the latest trunk, you can use the -u switch from x*set
> >  > >  to do the same:
> >  > >
> >  > >  xgroupset add n0000 -u (to add all groups)
> >  > >  xgroupset add n0000 -u (to add all users)
> >  > >
> >  > >  let us know if this works for you.
> >  > >  thanks!
> >  >
> >  > Well, xgroupset add -au worked fine, but when I did xuserset add -au
> >  > it bombed saying that the root user was already there:
> >  >
> >  > xuserset add n0001 -u
> >  > xp_user_add: n0001: Error 5: root:user exists
> >  > Error: root:user exists
> >  >
> >
> >
> > that is because you started xcpufs with the -u switch and it could add
> >  the "root" user to the pool since /root/.ssh/id_rsa.pub exists on the
> >  node (the perceus xcpu module creates that file before spawning xcpufs)
> 
> I tried it both ways.  I thought the -u could be the problem, but the
> same happens when I boot the nodes without it!
> 

it shouldn't do that unless the user already exists. i just checked and
yeah it does not skip over the users that it could not add. I will send
a fix soon.

thanks for hammering it out.


Reply via email to