I'm running PHP3 on an Intel box with Linux, Apache, and Oracle8i as the backend.
I've managed to compile PHP with Oracle support, but the performance is nowhere near the same implementation in mySQL. While Oracle is supposed to be slower but more scalable, and mySQL is FAST, I'm getting PHP parsing times a factor of 4 or 5 times more than mySQL. Nuts!
Aside from the OCI function calls, the code is for the most part the same, and the time it takes for Oracle to serve up the queries is so fast it doesn't even rate as a response time concern.
The delay seems to center around eregi functions, based on timers I set in the code. Any time this is called, parsing time is around 1 to two seconds, and if called 6 times, this gives a whopping 6 to 12 second delay.
Why is the performance of PHP with Oracle support so inferior to the performance of PHP-mySQL? Is there a binary distribution of an RPM out there I can use that's optimized, or a set of detailed steps to compile this sucker so it doesn't result in these horrendous performance times?
I've managed to compile PHP with Oracle support, but the performance is nowhere near the same implementation in mySQL. While Oracle is supposed to be slower but more scalable, and mySQL is FAST, I'm getting PHP parsing times a factor of 4 or 5 times more than mySQL. Nuts!
Aside from the OCI function calls, the code is for the most part the same, and the time it takes for Oracle to serve up the queries is so fast it doesn't even rate as a response time concern.
The delay seems to center around eregi functions, based on timers I set in the code. Any time this is called, parsing time is around 1 to two seconds, and if called 6 times, this gives a whopping 6 to 12 second delay.
Why is the performance of PHP with Oracle support so inferior to the performance of PHP-mySQL? Is there a binary distribution of an RPM out there I can use that's optimized, or a set of detailed steps to compile this sucker so it doesn't result in these horrendous performance times?