Re: [OpenAFS] Re: Couldn't get CPS for AnyUser, will try again in 30 seconds; code=5376.

[Brett Heroux]

That did it.

1) stop the ptserver
2) back up prdb files
3) remove prdb files
4) pt_util -user -group -members -name -system -prdb ./prdb.DB0 
-datafile /tmp/t

5) pt_util -user -group -members -name -system -w -datafile /tmp/t
6) start the ptserver
7) joy

I think the -system was unnecessary, but Thank You So Much Andrew.

Brett Heroux

On 4/7/2012 12:16 PM, Andrew Deason wrote:

On Fri, 6 Apr 2012 14:59:40 -0500
Brett Heroux  wrote:


The pt_util output looks good, it just gives all the users.

Still would appreciate help.

Would you be willing to provide your prdb.DB0? It just contains things
like usernames, groups, group memberships, ids, etc. It shouldn't
contain very sensitive information, unless any of your usernames or
group memberships etc are sensitive.

If you want something quicker to get stuff up and running, one thing
that will probably work is to recreate the ptdb. One way to do this is
to stop the ptserver, move the prdb.DB0 and prdb.DBSYS files out of the
way, and use pt_util to dump the information from them to a temporary
file. Then use pt_util to load that information into a new prdb.DB0, and
start up the ptserver again. Just make sure you back up
prdb.DB0/prdb.DBSYS

(See the pt_util documentation for that; I'm in a bit of a hurry to
provide more detailed info:
)



___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Ubuntu

[Russ Allbery]

Sergio Gelato  writes:

> I also suspect that it's going to take less effort to just maintain PPAs
> --- for me personally it's already quite clear, for the community at
> large it may be harder to estimate --- than to push updates through the
> Ubuntu process, but it doesn't seem right to surrender without even
> trying; so let's see what this experiment will lead to. I wouldn't
> single out Ubuntu:  Debian's own policies and practices can also be
> quite frustrating at times (and don't get me started on Apple's…)

Oh, yeah, there's nothing inherently wrong with the procedures given the
goals of the project, really.  Debian has other difficult procedures (such
as around licensing); people just tend not to notice them because I'm
actively involved in Debian and babysit OpenAFS through them.

-- 
Russ Allbery (r...@stanford.edu) 
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

[OpenAFS] Re: Couldn't get CPS for AnyUser, will try again in 30 seconds; code=5376.

[Andrew Deason]

On Fri, 6 Apr 2012 14:59:40 -0500
Brett Heroux  wrote:

> The pt_util output looks good, it just gives all the users.
> 
> Still would appreciate help.

Would you be willing to provide your prdb.DB0? It just contains things
like usernames, groups, group memberships, ids, etc. It shouldn't
contain very sensitive information, unless any of your usernames or
group memberships etc are sensitive.

If you want something quicker to get stuff up and running, one thing
that will probably work is to recreate the ptdb. One way to do this is
to stop the ptserver, move the prdb.DB0 and prdb.DBSYS files out of the
way, and use pt_util to dump the information from them to a temporary
file. Then use pt_util to load that information into a new prdb.DB0, and
start up the ptserver again. Just make sure you back up
prdb.DB0/prdb.DBSYS

(See the pt_util documentation for that; I'm in a bit of a hurry to
provide more detailed info:
)

-- 
Andrew Deason
adea...@sinenomine.net

___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

[OpenAFS] RT Access was Feature wish: remove partition while fileserver keeps on running

[Jeffrey Altman]

On 4/7/2012 11:20 AM, Jason Edgecombe wrote:
>> Done. We really should sort out RT so that you, and others, have more
>> access to the bug database.
>>
>> S.
>>
> Thanks!
> 
> Who can make that change?
> 
> Jason

there are several subjects raised by this thread.

1. Who should have closed RT130561?

The gatekeeper that merged the change that "FIXED" the ticket but only
after the gerrit ticket and the commit sha1 were added as comments to
the RT ticket AND only if the change was not going to be pulled up to
another branch.

2. Who runs OpenAFS RT?

OpenAFS relies upon the central.org RT for services.  We do this so that
the OpenAFS gatekeepers do not need to be sysadmin for RT.  central.org
is maintained by Jeffrey Hutzelman and Chaskiel Grundman.

3. Who should have RT accounts?

In theory, any legitimate user that wants to submit a bug report.
This has proven hard to do in the past because an open account
registration process previous resulted in many false accounts being
created that were in turn used to deface RT.  Not just for OpenAFS but
for the other services that rely on the central.org RT.

There is certainly a broader list of active community members that could
be given accounts.  A list needs to be generated and given to the RT
administrator.

4. What privileges should individuals that have accounts be granted?

The ability to resolve tickets is at this point limited to the
gatekeepers and a small number of individuals that have a track record
of fixing bugs that are reported to RT.

I don't know the specific privileges that each individual account holder
has because I am not an administrator for RT and do not have privileges
to make changes.

Jeffrey Altman



signature.asc
Description: OpenPGP digital signature

Re: [OpenAFS] Feature wish: remove partition while fileserver keeps on running

[Jason Edgecombe]

On 04/07/2012 11:10 AM, Simon Wilkinson wrote:

On 7 Apr 2012, at 15:50, Jason Edgecombe wrote:


Gerrit commit<  http://gerrit.openafs.org/7108>  includes the previously 
discussed /vicepXX/NeverAttach flag. The commit has been merged into the master 
branch.

I can't close ticket 130561. Would someone close it, please?

Done. We really should sort out RT so that you, and others, have more access to 
the bug database.

S.


Thanks!

Who can make that change?

Jason
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Feature wish: remove partition while fileserver keeps on running

[Simon Wilkinson]

On 7 Apr 2012, at 15:50, Jason Edgecombe wrote:

> Gerrit commit< http://gerrit.openafs.org/7108 > includes the previously 
> discussed /vicepXX/NeverAttach flag. The commit has been merged into the 
> master branch.
> 
> I can't close ticket 130561. Would someone close it, please?

Done. We really should sort out RT so that you, and others, have more access to 
the bug database.

S.

___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Feature wish: remove partition while fileserver keeps on running

[Jason Edgecombe]

Gerrit commit< http://gerrit.openafs.org/7108 > includes the previously 
discussed /vicepXX/NeverAttach flag. The commit has been merged into the 
master branch.


I can't close ticket 130561. Would someone close it, please?

https://rt.central.org/rt/Ticket/Display.html?id=130561

Thanks,
Jason

On 03/01/2012 08:31 PM, Jason Edgecombe wrote:

Works for me.

On 02/29/2012 11:49 PM, Derrick Brashear wrote:

Given we have "AlwaysAttach", "NeverAttach" would seem to be the
obvious opposite.

On Wed, Feb 29, 2012 at 8:13 PM, Andrew 
Deason  wrote:

On Mon, 27 Feb 2012 18:18:38 -0500
Jason Edgecombe  wrote:


What I think would be an even more useful feature is the ability to
detach and reattach a partition on another server (via SAN storage or
whatever), since there aren't very good workarounds for that. If we
could do that, we could also just detach a partition and reattach 
it to

the same server again.

May I suggest an alternate approach?

In http://rt.central.org/rt/Ticket/Display.html?id=130561
I documented an idea about a "DoNotAttach" flag. While this would not
allow for hot adds and hot removals, it would allow you to flag a
/vicepX folder for removal. After that, just schedule a service 
restart.

Yeah, that sounds useful and easy to do, to me. I don't entirely
consider it an alternate approach in that it is a full replacement for
what I wrote above, but it would certainly be useful.

--
Andrew Deason
adea...@sinenomine.net

___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info





___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info



___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Ubuntu

[Sergio Gelato]

* Harald Barth [2012-04-07 11:09:53 +0200]:
> 
> > I'm doing this for my site.
> 
> As in "su" or in "astro"?

As in "astro", for the time being. If others within "su" want to leverage 
my work that can be arranged, but so far I haven't been asked.

> > It also has the duplicate NAT ping issue, which can cause clients to flood
> > file servers with substantial amounts of useless traffic in some
> > situations.
> 
> Which in itself should be enough to stop it from being released. But
> I'm pessimistic, Ubuntu has in spite of its popularity issues when it
> comes to fixing bugs that they can dismiss by blaming upstream debian.
> But maybe my experience is biased from the messy state the heimdal
> package is in.

Well, I've just filed two bug reports, one for the NAT ping issue and one
for the fileserver issue. I may try to cover the idledead issues later
today. (I need to research them a bit first.)

I also suspect that it's going to take less effort to just maintain PPAs
--- for me personally it's already quite clear, for the community at large 
it may be harder to estimate --- than to push updates through the Ubuntu 
process, but it doesn't seem right to surrender without even trying; so
let's see what this experiment will lead to. I wouldn't single out Ubuntu: 
Debian's own policies and practices can also be quite frustrating at times
(and don't get me started on Apple's…)
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Ubuntu

[Russ Allbery]

Harald Barth  writes:

> Which in itself should be enough to stop it from being released. But I'm
> pessimistic, Ubuntu has in spite of its popularity issues when it comes
> to fixing bugs that they can dismiss by blaming upstream debian.  But
> maybe my experience is biased from the messy state the heimdal package
> is in.

Heimdal tends to get caught in the same situation that OpenAFS is in,
where there's no one in Ubuntu specifically paying attention and the
Debian maintainer isn't really paying attention to Ubuntu.  It looks like
this may be changing; the newly active Debian maintainer seems to be doing
bug triage in Ubuntu as well.

-- 
Russ Allbery (r...@stanford.edu) 
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Ubuntu

[Harald Barth]

> I'm doing this for my site.

As in "su" or in "astro"?

> It also has the duplicate NAT ping issue, which can cause clients to flood
> file servers with substantial amounts of useless traffic in some
> situations.

Which in itself should be enough to stop it from being released. But
I'm pessimistic, Ubuntu has in spite of its popularity issues when it
comes to fixing bugs that they can dismiss by blaming upstream debian.
But maybe my experience is biased from the messy state the heimdal
package is in.

Harald.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info