Those commands allowed me to remove the experiment. Thank you.
I'm fairly certain that the error was caused by a problem in dhcpd.conf
Each time I add a new node, when dhcpd.conf is automatically re-written, the allowed MAC addresses are changed to an incorrect value. I had added a new node just before running this experiment, and neglected to manually fix dhcpd.conf
Regards,
T Ficarra
--- On Wed, 7/8/09, Leigh Stoller <stoller@flux.utah.edu> wrote:
From: Leigh Stoller <stoller@flux.utah.edu>
Subject: Re: [Testbed-admins] Deleting an Experiment
To: "Timothy Ficarra" <t.ficarra@yahoo.com>
Cc: testbed-admins@flux.utah.edu
Date: Wednesday, July 8, 2009, 5:51 PM
> Here is the output from those commands:
Okay, so the thing to do is:
mysql> update experiment_stats set lastrsrc=null where exptidx='28';
mysql> update experiments set state='active',expt_locked=null,
mysql> canceled=0 where idx=28;
This will make the experiment appear swapped in, so go ahead and swap
it out.
The reason for doing this is that I do not know how far into the swap
process it got, so this will clean up any other swap state that might
have been created.
It would be good to know why the activation got stuck. Did you boss
reboot or did the log show any errors?
Lbs
|