We have had a MySQL/Apache/Perl combo running fairly smoothly. However the other day MySQL seemingly went crazy, and brought the rest of the system down; no sockets were available, even to ssh into. After a few minutes, enough sockets were available to ssh into, however MySQL seemed to far gone, and had to be restarted. I have a feeling it is due to the scripts we are running (mod_perl via apache and regular perl, via local cronjobs) and the way they are written. All the scripts connect to MySQL via local UNIX domain sockets, thus creating a /tmp/mysql.sock for each invocation. Is this the best way to do it, or should TCP/IP sockets be used, by connecting to the full hostname, versus connecting to localhost, which seems to use local sockets. I know the local sockets are generally considered faster, however I have also heard that the Linux kernel does not handle a lot of local sockets well. Would a combination of local and IP sockets be used, or strictly IP, since strictly local sockets seems to bring the system to the edge. As much information, links, comparasions, benchmarks, data on the subject would be greatly appreciated. Don't hesitate to contact me for additional questions/info. Thank you, Hans Zaunere [EMAIL PROTECTED]
--------------------------------------------------------------------- Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail <[EMAIL PROTECTED]> To unsubscribe, e-mail <[EMAIL PROTECTED]> Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
