Print

Print


  Hi Giuseppe,

On Sun, Feb 27, 2005 at 07:47:01PM +0100, Giuseppe Della Ricca wrote:
> On Sun, 27 Feb 2005, Peter Elmer wrote:
> >   Thanks for the bug report. I've CC-d it to the xrootd mailing list. (I 
> > think, but am not 100% sure, that you can now post to the list w/o being 
> > subscribed...)
> 
> ok, let's see if it works ...

  I got the mail, but I was explicitly on the To:....

> >   Andy, see below (there is a missing else). I'll let you fix it...
> > 
> >   Giuseppe, does xrootd refuse to start altogether after this error or does 
> > the message only appear in the log file?
> 
> xrootd refuses to start

  Ok, good.

> >   Also, are you setting this up for analysis use or for simulation production?
> 
> for simulation production at INFN in Rome
> 
> We've been using xrootd since the beginning of SP6, to access the 
> background triggers (we have had too many problems with NFS in the past).
> 
> (and this setting is important for security)
> 
> Now I'm setting up SP7, and while I was fighting with KanAccess.cfg 
> ("configuration error: path has to be of the form <server>:<port>/"),
> I updated the xrootd version and found this bug.

<babar-specific-comment>
  Ah, yes, originally you could put things like "<server>:<port>/dir" in
the KanAccess.cfg file when specifying xrootd for use. This duplicates
(and exposes to the user) the equivalent of the "localroot" functionality of 
xrootd itself, so we eliminated the option in KanAccess.cfg.
</babar-specific-comment>

                                   Pete

-------------------------------------------------------------------------
Peter Elmer     E-mail: [log in to unmask]      Phone: +41 (22) 767-4644
Address: CERN Division PPE, Bat. 32 2C-14, CH-1211 Geneva 23, Switzerland
-------------------------------------------------------------------------