Re: 503 service not available when a fastcgi dies

On Sat, Dec 04, 2004 at 09:30:46AM -0500, Kevin Watt wrote:
> If one of my ruby fastcgi processes dies, lighttpd refuses to process any
> more requests, giving 503 service not available.
> 
> Sometimes it works a few times before this happens, I suspect it just
> doesn't move on to the next round-robin before it gives the error.  I've
> also noticed it using the first child only for hte first 5 minutes (strace
> on it shows it handling all the requests, while the others are idle).
> Eventually the load spreads out though.  Not sure what could be causing
> that.
> 
> I'd be preferable that it drop that socket from the list + continue to use
> the 4 others that are still fine.  Even better would be a way to have it
> automatically respawn.  Any ideas?

If I could reproduce it would be easy to fix for me. 

What shall I do to set something up like your environment ?

> Ciao,
> Kevin
> 
    Jan

--

-- 
Jan Kneschke                                     http://jan.kneschke.de/
Perhaps you want to say 'thank you, jan':    http://jk.123.org/wishlist/