[OpenAFS] PTS problems, serious?

Jeffrey Hutzelman jhutz@cmu.edu
Sun, 28 May 2006 14:03:37 -0400


On Thursday, May 18, 2006 05:45:38 PM +0300 Juha J=E4ykk=E4 <juolja@utu.fi> =

wrote:

> Hi!
>
> I just ran into this:
>
> ~> pts createuser foo -id 42
> pts: major synchronization error ; unable to create user foo
>
> What does this mean? How do I fix this? Or, is there anything to fix,
> since
>
> ~> pts listentries
> ...
> foo		42	-204	2
>
> How serious a problem do I have? To me, this sounds like ubik has some
> problems!

This error indicates that some server is confused about what is the current =

transaction or who is the sync site.  In normal operation with correctly=20
configured servers, this should be quite rare and always transient.

If you are able to read back the change, then it made it to at least one=20
server, and the confused server will have been marked out of date and will=20
eventually pick up a new database.  So, there is no long-term ill effect.

However, if you are seeing this on a regular basis, then it seems likely=20
you have some sort of configuration problem.  Make sure all of your=20
dbservers have the _same_ server-side CellServDB, and that it lists each=20
server exactly once.  If the servers have more than one interface, only one =

interface per server should be listed in the server-side CellServDB.

-- Jeffrey T. Hutzelman (N3NHS) <jhutz+@cmu.edu>
   Sr. Research Systems Programmer
   School of Computer Science - Research Computing Facility
   Carnegie Mellon University - Pittsburgh, PA