[OpenAFS-devel] Fwd: [grand.central.org #128372] AutoReply: Output from 'fs getfid'
doesn't complain about paths not in /afs
Phillip Moore
w.phillip.moore@gmail.com
Mon, 18 Oct 2010 14:45:48 -0400
--0015175caa24b45c0f0492e8956b
Content-Type: text/plain; charset=ISO-8859-1
I just submitted this bug report, and then found a much more serious problem
with getfid -- it ALWAYS exits 0. I'd like to comment on this bug report,
rather than create a new one, since I would probably try to fix both
problems with one patch.
However, I don't have a login for the RT site. How do I get one?
Also, what's the policy/preference here? Strictly speaking, there are two
issues here (no feedback on paths that *should* generate errors, and then
the fact that is no error code at all). I'm lazy -- I'd just fix these in
one patch, but does the dev community prefer that these be addressed
separetly?
Am I thinking too hard?
---------- Forwarded message ----------
From: OpenAFS Bug Reports via RT <openafs-bugs@openafs.org>
Date: Mon, Oct 18, 2010 at 2:32 PM
Subject: [grand.central.org #128372] AutoReply: Output from 'fs getfid'
doesn't complain about paths not in /afs
To: w.phillip.moore@gmail.com
Greetings,
This message has been automatically generated in response to the
creation of a trouble ticket regarding:
"Output from 'fs getfid' doesn't complain about paths not in /afs",
a summary of which appears below.
There is no need to reply to this message right now. Your ticket has been
assigned an ID of [grand.central.org #128372].
Please include the string:
[grand.central.org #128372]
in the subject line of all future correspondence about this issue. To do so,
you may reply to this message.
Thank you,
openafs-bugs@openafs.org
-------------------------------------------------------------------------
The fs getfid command (wish I had this one 10 years ago...) behaves very
differently from all the other fs commands that accept a -path argument.
If you pass a list of paths, and some of them are not in AFS, or aren't
present, etc. the command gives you feedback on this:
[root@rpcore ~]# fs listacl /afs/a.rp.nyc.us.test.efs/home/efsops /foo
/var/tmp
Access list for /afs/a.rp.nyc.us.test.efs/home/efsops is
Normal rights:
system:administrators rlidwka
system:anyuser rl
efsops rlidwka
fs: File '/foo' doesn't exist
fs: Invalid argument; it is possible that /var/tmp is not in AFS.
This is true for all of the following: diskfree, examine, getcalleraccess,
listacl, listquota, quota, storebehind, whereis, and whichcell.
But not getfid. There's no valid reason, IM!HO, that getfid's feedback
should not be consistent with the other commands.
--0015175caa24b45c0f0492e8956b
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
<div><br></div>I just submitted this bug report, and then found a much more=
serious problem with getfid -- it ALWAYS exits 0. =A0 I'd like to comm=
ent on this bug report, rather than create a new one, since I would probabl=
y try to fix both problems with one patch.<div>
<br></div><div>However, I don't have a login for the RT site. =A0How do=
I get one?</div><div><br></div><div>Also, what's the policy/preference=
here? =A0 Strictly speaking, there are two issues here (no feedback on pat=
hs that *should* generate errors, and then the fact that is no error code a=
t all). =A0 I'm lazy -- I'd just fix these in one patch, but does t=
he dev community prefer that these be addressed separetly? =A0=A0</div>
<div><br></div><div>Am I thinking too hard?<br><br><div class=3D"gmail_quot=
e">---------- Forwarded message ----------<br>From: <b class=3D"gmail_sende=
rname">OpenAFS Bug Reports via RT</b> <span dir=3D"ltr"><<a href=3D"mail=
to:openafs-bugs@openafs.org">openafs-bugs@openafs.org</a>></span><br>
Date: Mon, Oct 18, 2010 at 2:32 PM<br>Subject: [<a href=3D"http://grand.cen=
tral.org">grand.central.org</a> #128372] AutoReply: Output from 'fs get=
fid' doesn't complain about paths not in /afs<br>To: <a href=3D"mai=
lto:w.phillip.moore@gmail.com">w.phillip.moore@gmail.com</a><br>
<br><br><br>
Greetings,<br>
<br>
This message has been automatically generated in response to the<br>
creation of a trouble ticket regarding:<br>
=A0 =A0 =A0 =A0"Output from 'fs getfid' doesn't complain =
about paths not in /afs",<br>
a summary of which appears below.<br>
<br>
There is no need to reply to this message right now. =A0Your ticket has bee=
n<br>
assigned an ID of [<a href=3D"http://grand.central.org" target=3D"_blank">g=
rand.central.org</a> #128372].<br>
<br>
Please include the string:<br>
<br>
=A0 =A0 =A0 =A0 [<a href=3D"http://grand.central.org" target=3D"_blank">gr=
and.central.org</a> #128372]<br>
<br>
in the subject line of all future correspondence about this issue. To do so=
,<br>
you may reply to this message.<br>
<br>
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0Thank you,<br>
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0<a href=3D"mailto:openafs-b=
ugs@openafs.org">openafs-bugs@openafs.org</a><br>
<br>
-------------------------------------------------------------------------<b=
r>
The fs getfid command (wish I had this one 10 years ago...) behaves very<br=
>
differently from all the other fs commands that accept a -path argument.<br=
>
If you pass a list of paths, and some of them are not in AFS, or aren't=
<br>
present, etc. the command gives you feedback on this:<br>
<br>
[root@rpcore ~]# fs listacl /afs/a.rp.nyc.us.test.efs/home/efsops /foo<br>
/var/tmp<br>
Access list for /afs/a.rp.nyc.us.test.efs/home/efsops is<br>
Normal rights:<br>
=A0system:administrators rlidwka<br>
=A0system:anyuser rl<br>
=A0efsops rlidwka<br>
<br>
fs: File '/foo' doesn't exist<br>
fs: Invalid argument; it is possible that /var/tmp is not in AFS.<br>
<br>
This is true for all of the following: diskfree, examine, getcalleraccess,<=
br>
listacl, listquota, quota, storebehind, whereis, and whichcell.<br>
<br>
But not getfid. =A0There's no valid reason, IM!HO, that getfid's fe=
edback<br>
should not be consistent with the other commands.<br>
<br>
</div><br></div>
--0015175caa24b45c0f0492e8956b--