Re: [Sks-devel] recon segfaulting

[Yaron M. Minsky]

If you built the database with the bad patch, then that was probably the
problem.  Removing the patch doesn't undo the database corruption.  Was
the problematic build run with the patch in question?

y

On Wed, 2004-03-17 at 12:35, Jan Dreyer wrote:
> Hello,
> 
> Peter Palfrader wrote:
> 
> > Try to build sks without the slong_dlong patch.  i.e. checkout the svn,
> > and rm debian/patches/509_Slong_Dlong.  Please report if this fixes your
> > problem.
> 
> No, it doesn't.
> 
> Removing my database and starting with a (nearly) empty one does it (same
> with or without slong_dlong), but there is no sense in an empty keyserver,
> isn't it?
> 
> So I suppose the database is still corrupted.
> Where or how do I get a safe database?
> 
> I'm running now the small-DB-Version; that may cause some more gossip with
> my partners ...
> 
> ------- UPDATE ---------
> I tried to do it the long, long way:
> I merged a dumpfile, started the server, watched for the recon-process,
> stopped the server, merged the next and so on.
> 
> It seems to me that worked! Our Server is now up and syncing.
> Maybe the recon-process doesn't handle too much hash-updates?
> 
> Greetings
> Jan Dreyer
-- 
|--------/            Yaron M. Minsky              \--------|
|--------\ http://www.cs.cornell.edu/home/yminsky/ /--------|

Open PGP --- KeyID B1FFD916
Fingerprint: 5BF6 83E1 0CE3 1043 95D8 F8D5 9F12 B3A9 B1FF D916