Here's our scenario: Ubuntu 10.10 x64 server install. Default PHP installation, latest Cherokee from SVN.
Simple low-load, low use PHP scripts running. Almost unused at this point. Have multiple sources set up for PHP within Cherokee via FastCGI Handler: Backends: PHP Interpreter 127.0.0.1:47990 php local 2 127.0.0.1:47991 php local 3 127.0.0.1:47992 php local 4 127.0.0.1:47993 php local 5 127.0.0.1:47994 Have those set up in failover mode within Cherokee. One or more of these backends becomes non responsive (process still running though). No clue on my part why any such process is failing at this point. Let's say 47990 is in this failed mode, I would assume Cherokee failover should be able to detect this in timely manner and ask 47991 to handle the request. Well this doesn't happen. The connection sits there waiting and waiting until the timeout defined for this server (10 seconds). So two questions: 1. Any ideas on a fix or cause for PHP failure like this? What should I be looking at to debug PHP's issues like this? 2. Am I hoping for too much from the failover functionality at this time (i.e. a configurable wait time for failover detection and cascading attempts down the list)? Should I submit a Ticket for Request for Enhancement? _______________________________________________ Cherokee mailing list [email protected] http://lists.octality.com/listinfo/cherokee
