I wanted to upgrade the aegir tools on the -dev and -prod sites. I ran into some problems running the upgrade steps; which are probably due to some mistake I had made in managing the changes made since the original install. During the upgrade process, I was getting a variety of database error messages. I'm assuming that part of the problem was that I had originally installed 0.2, then upgraded to 0.3 and somewhere along the line, upgraded drush. Another symptom I was seeing was that my crontab for aegir was gone, so I could not process tasks. I was also seeing some issues with permissions on writing out the backup taken during the migration process (I think).
I thought about continuing on with the upgrade process, but backed out the changes (do follow the upgrade instructions, so you've got good backups), got stable again and then quit for the night.
I thought about this overnight and decided that I might be better served with a re-install. So, I took some more backups and then began a re-install of 0.4a5. Worked perfectly.
For my purposes, dropping the aegir database and recreating it, migrating in the various platforms, sites, web and database servers posed little risk to my operations. What I got out of this approach was running the new install.sh script and seeing how drush_make pulled down all the dependencies and some minty green freshness as all the verifications went green!
Attached to this post is the shell output where I was getting errors.