Authors can't create new wiki page or save edits

That sounds like a Server Error 500 to me. You may be running up against a PHP memory limit or something.

Try installing TPC Memory Usage to see how much memory you have and if you’re peaking the CPU resources.

https://www.google.com/search?q=server+error+500&aq=f&oq=server+error+500

I contacted our server administrator who says the following: For some reason, the PHP script trying to create/edit a new post uses all of the allowed memory (128MB in our case).
We could increase that limit, however it is a global server setting and it’s already set pretty high, so I’d like to try and find out first why that script would use so much memory.

The server admin could see a bunch of fatal errors in the log as the allowed memory size was exhausted. If you have time, could you clarify why we would need more memory?

Thanks for your help.
Pam

Can you copy the exact error messages that are getting logged and maxing out your PHP limit? I’ll do some checking on my end but there’s no reason creating a wiki article would use any more memory than any other post type. I’ll look into it. Any info you can dig up would be helpful.

Thanks!

Here’s the error log – thanks for taking a look.

[Thu Mar 28 13:55:21 2013] [error] [client 127.0.0.1] PHP Fatal error: Allowed memory size of 134217728 bytes exhausted (tried to allocate 71 bytes) in /opt/wordpress/commons.yale.edu/releases/20130315-1925/wp-includes/wp-db.php on line 1449, referer: https://rhythmcoglab.commons.yale.edu/wp-admin/post.php?post=1368&action=edit

More clarification from the server admin: These messages are from the Apache server and they don’t have user information. The 128M is a global parameter that will kill any PHP script that tries to use up more than that. Obviously, there’s something different between the different users since some of them always work and others always fail. I was hoping the developer could help us understand exactly why this is happening (as you mentioned, it could be large cache, revision history, etc) and if there’s something we can do in the app to fix it. If they believe this is normal behavior, I can increase the limit to 192 or 256M.

My question to the server admin was – why are we focusing on the wiki plugin, this seems like it might have more to do with the multi-site instance and a users number of posts/comments/revisions in the database over the multi-site as a whole. What do you think?

There could be a million reasons you’re hitting your PHP limit and you’re right, multisite instances — in my experience — tend to be more resource-intensive than non-multisite instances. As I said, I don’t see any specific reason why the wiki plugin would do this more than any other plugin or post type — there is nothing unique about saving/publishing a wiki article that’s any different from saving/publishing a regular WordPress post or page.

I just tested locally on a multisite installation and nothing jumped out at me. There weren’t any major errors or warnings that were being thrown in the code.

Some things to look at specifically since you’re running multisite are the plugins you’re using particularly those that are network active. Not all plugins need to be network activated and network activating plugins may have an adverse effect on the overall performance of the site since you are requiring those scripts to be run on every page load. If there are things that you don’t need to run on every single sub-site, I would recommend deactivating them across the network and just activating them on those sites that need them. The TPC Memory Usage plugin I mentioned before will also tell you which functions/scripts are responsible for the most load, so that might help in tracking down what’s actually causing the problem.

I’m guessing it’s one of those situations where you’re hovering right on the border of maxing out your memory and the post action is just the straw that breaks the camel’s back rather than it being specifically related to the wpwtds plugin, but I’m certainly open to trying to optimize the plugin if it turns out that there are things in the code that could or should be done better I’m just not seeing anything that would directly lead to a memory leak type situation in my plugin.

Chris, thank you for your time. I’m requesting they bump up the memory allocation to see if that solves the issue. I will try the TPC Memory Usage plugin as well.

No problem. I agree with your server admin that 128M is usually more than enough. It would be helpful to know what scripts or functions are leading to the server using more memory but bumping up the memory allocation couldn’t hurt anything and would ultimately only benefit your site since you’re running multisite, especially if you’re planning on having a lot of sub-sites.