Reconnecting after getting a fatal "100" error

October 30, 2003

About one week ago our user database crashed. No data was lost, but
there was a brief period of time where the server was in such an odd
state that it accepted current results from users, but then corrupted
their local user_info.sah file, which contains information about user
id and credit.

Days later, when everything was normal, users started to return
results infected with the bogus data in their user_info.sah file.
Unfortunately our server rejects these particular results with a
cryptic, fatal "100" error.

The problem is, users are still left with bogus user_info.sah files,
and their result.sah files remain infected as well. So the only
trustworthy solution is to just remove these offending files (search
for the files "result.sah" and "user_info.sah" on your systems and
delete them). When you start up SETI@home it will ask you to log in
again. The process of logging in with clean up the user_info.sah
files, and download a fresh workunit to chew on.

However, this also nullifies any work done on the infected result. We
apologize for that wasted effort, but in the interest of clean science
the best thing to do is just remove these files and start again on a
new workunit from scratch.

People using third-party "add-on" software to cache units may be
experiencing a set of other problems that our beyond our jurisdiction,
since we haven't written that software and don't know how it works or
why it is broken. There are several threads on our message boards
discussing potential solutions depending on which add-on you are
using.

Return to SETI@home Page
Copyright ©2001 SETI@home 


Copied from this locaton:
http://setiathome.ssl.berkeley.edu/100error.html