Thread

Normally, the profiler indicates that about 3M of memory is needed to process the templates for the frontpage of our site, but twice in the last couple of weeks, the EE page has gone down with an out of memory PHP error, for every request until it has enough memory.  We’ve bumped the memory up to 256M (then 512M the second time) and restart apache and the frontpage will then load, and be back down to 3M on the average load.

Other sections of the site load just fine, it seems to be limited at the single default template at the front of the site.

What I am wondering is: Is there perhaps regular cleanings/cache and database tasks that occasionally get called on page request that could cause these one-time huge bumps in memory until they are complete?  I’m imagining there is something like WP’s WP-Cron functions. How do I disable that requirement, if that is what is happening, and set those funcitons to run on cron instead?

Thanks

M.

Also, we’ve had one event that the exp_sessions table has gotten corrupt.  Is there any reason not to alter the exp_sessions table to innodb?  I find that innodb is more tolerant to massive simultaneous read/writes from web clients.

Is there caching being done on the home page template?

Have you turned on the output profiler to see what’s using the memory?

Yes, the profiler shows 3-5M of memory usage very consistently.

The homepage does have an a couple sections with cache=“yes” refresh=“60” dynamic=“no” turned on.

In the Template options, Enable Caching was set to “Yes” and Refresh Interval (In minutes) was set to 0.  I just set “Enable Caching” to “No”.