Anaconda failure on i686 - lvcreate fails for swap

[G.Wolfe Woodbury]

These are rough times for anaconda, first one failure, then another and 
another -- it's driving me bonkers.

The latest failure is anaconda failing silently (no traceback) on an LV 
create of the swap filesystem in a "pure vanilla" install.

No traceback means no bug report to make.

[Clyde E. Kunkel]

On 03/27/2009 08:56 PM, G.Wolfe Woodbury wrote:
> These are rough times for anaconda, first one failure, then another and
> another -- it's driving me bonkers.
>
> The latest failure is anaconda failing silently (no traceback) on an LV
> create of the swap filesystem in a "pure vanilla" install.
>
> No traceback means no bug report to make.

FWIW, I noticed that anaconda appeared to be hung without traceback, but 
it took off 40-45 minutes later and installed.  I had been assuming it 
was broken and killing the install, but it is doing something, then 
going on.  The developers are aware.

[G.Wolfe Woodbury]

Clyde E. Kunkel wrote:
> On 03/27/2009 08:56 PM, G.Wolfe Woodbury wrote:
> > These are rough times for anaconda, first one failure, then another and
> > another -- it's driving me bonkers.
> >
> > The latest failure is anaconda failing silently (no traceback) on an LV
> > create of the swap filesystem in a "pure vanilla" install.
> >
> > No traceback means no bug report to make.
>
> FWIW, I noticed that anaconda appeared to be hung without traceback, but 
> it took off 40-45 minutes later and installed.  I had been assuming it 
> was broken and killing the install, but it is doing something, then 
> going on.  The developers are aware.

In my case, it crashed and vt01 was ready to reboot.  Examination of the 
alternate windows showed nothing to give a clue of what went wrong.

[David Cantrell]

On 03/27/2009 02:56 PM, G.Wolfe Woodbury wrote:
> These are rough times for anaconda, first one failure, then another and
> another -- it's driving me bonkers.
>
> The latest failure is anaconda failing silently (no traceback) on an LV
> create of the swap filesystem in a "pure vanilla" install.
>
> No traceback means no bug report to make.

If you haven't yet filed a bug, please do at bugzilla.redhat.com.  For 
those who do not know, anaconda for Fedora 11 is getting the entire 
storage backend rewritten.  That's the code that deals with disks, 
drives, partitions, raid, lvm, filesystems, and so on.  Details:

https://fedoraproject.org/wiki/Features/AnacondaStorageRewrite

So please file bugs.  Even if there is no traceback, file a bug and 
describe what happened, including what steps you took during 
installation.  Also include a description of your hardware so we can 
reproduce the problem locally.

Pretty much all of us on the anaconda team have been working day and 
night with few breaks to get this code rewritten.

Several people have asked why we are rewriting the storage code in 
anaconda.  It's the last major piece of anaconda that's still mostly 
original anaconda code.  April 24, 2009 is the 10 year anniversary of 
anaconda.  A lot has changed in the world of storage in 10 years and we 
need code that can scale and adapt as necessary.

[cornel panceac]

2009/3/29 David Cantrell
> Pretty much all of us on the anaconda team have been working day and night
> with few breaks to get this code rewritten.
>
> thank you.
> --
> David Cantrell 
> Red Hat / Honolulu, HI
>
> --
> fedora-test-list mailing list
> [e-mail]
> To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list