Question about MUPDATE master mailboxes list

Rob Siemborski rjs3 at
Tue Feb 17 16:26:38 EST 2004

On Tue, 17 Feb 2004, Etienne Goyer wrote:

> Thanks, that answered most of my questions.  One last thing :
> On Tue, Feb 17, 2004 at 03:43:09PM -0500, Rob Siemborski wrote:
> > It mostly should depend on how much of the database is already populated
> > in the mupdate master.  If there's nothing there, and the database is
> > syncing on every write, then I would expect 500,000 entries to take a VERY
> > long time to run.
> Is it possible to disable the syncing on every write ?  I don't mean to
> use that in production, but when rebuilding database from scratch I
> guess it would speed up thing significantly.

In 2.2 there is the skiplist_unsafe config option (and an enviornment
variable, in 2.1).  If you're using Berkeley, you can set your database
backend to db3_nosync or bdb_nosync.

> Any idea about how to find what have caused database corruption ?  I
> kept a copy of the malfunctionning mailboxes.db before I rebuilt it.
> I'd really like to understand what went wrong in case I might do
> something to prevent this kind of situation in the future.

I'm sure its possible.  I'm not sure when I'd ever have the time unless
its a really obvious problem (and given that we're not seeing it
regularly, I'd guess it isn't).


Rob Siemborski * Andrew Systems Group * Cyert Hall 207 * 412-268-7456
Research Systems Programmer * /usr/contributed Gatekeeper

Home Page:
List Archives/Info:

More information about the Info-cyrus mailing list