Hi Ricardo, > I didn't know that the lag for a > report to get published could be of several hours. But that's because > I thought that the publishing process was totally automated. Is that > so, or is there some human intervention involved to publish the PASS / > FAIL reports in the cpantesters.org web site and the stats site? Or is > the lag just due to a very large queue?
Currently the reports are posted via email to [email protected] mailing list, which are then passed through to the NNTP server. The CPAN Testers server then polls the NNTP server every 4 hours (will soon be moving to hourly) and starts processing the reports into the cpanstats database. There is then a page builder which runs constantly, and runs based on a weighting system building the author and distribution pages as required. Typically there is at least a 6 hour delay between a report being posted and being listed on the cpantesters.org website. It is occasionally quicker, but with the high volume of reports being processed (20k+ a month) it does take time. The builder gets through roughly 400 pages an hour, and with the move to an hourly update, it should reduce the lag somewhat. CT2.0 should improve the lag between email to NNTP too. Must get round to writing a talk about all this ;) Cheers, Barbie. -- Birmingham Perl Mongers <http://birmingham.pm.org> Memoirs Of A Roadie <http://barbie.missbarbell.co.uk> CPAN Testers Blog <http://blog.cpantesters.org>
