Edit report at http://bugs.php.net/bug.php?id=53857&edit=1
ID: 53857 User updated by: samfiragabriel at gmail dot com Reported by: samfiragabriel at gmail dot com Summary: Pool definition errors Status: Bogus Type: Feature/Change Request Package: FPM related Operating System: GNU/Linux PHP Version: 5.3.5 Block user comment: N Private report: N New Comment: Hello, I apologise for the late response, i was on vacation for the past week. Also, sorry if this was not the right way to go about it. I know it is not really a bug, that's why I selected "Feature/Change Request". I was taking a look at fpm_conf_process_all_pools() and it just seamed like a good place to remove a pool from fpm_worker_all_pools if there is a configuration error for that pool. But if its a design decision that cannot be changed, then the --test flag is ok. It just seamed like it could be useful to log a configuration error, skip the pool, and continue without starting only that pool. Thank you for your time and for answering to my post! Best regards, Gabriel Previous Comments: ------------------------------------------------------------------------ [2011-01-29 11:22:51] f...@php.net Thank you for taking the time to write to us, but this is not a bug. Please double-check the documentation available at http://www.php.net/manual/ and the instructions on how to report a bug at http://bugs.php.net/how-to-report.php If one of your configuration line is wrong, it's detected and FPM won't start. Here is the example you mentioned (non existant chroot dir) (in this example /nowhere does not exist) # grep ^chroot /path/to/php-fpm.conf chroot = /nowhere # ./sapi/fpm/php-fpm [29-Jan-2011 11:16:34] ERROR: [pool www_chroot] the chroot path '/nowhere' does not exist or is not a directory [29-Jan-2011 11:16:34] ERROR: failed to post process the configuration If you run php-fpm with --test, you can anticipate an configuration error and you won't restart FPM. If you have a practical case which is not detected by the --test, please open a bug report and it'll be corrected. thx ------------------------------------------------------------------------ [2011-01-27 17:08:30] samfiragabriel at gmail dot com Description: ------------ It seams that if there is an error in the php-fpm config file, lets say in one of the pools, then fpm just dies with an error. For example, if you define a chroot value in one of the pools, and the directory does not exist, the master process and all the other pools just die upon SIGUSR2. In a shared hosting environment much of the configuring is done via scripts, and rarely by hand. If one value is erroneous, it would be better if that particular pool is skipped instead of just dying all together. That way at least only one site will go down instead of all of them. ------------------------------------------------------------------------ -- Edit this bug report at http://bugs.php.net/bug.php?id=53857&edit=1
