Hey Folks, thanks in advance for your help.

I realize this topic may be vague, and if even if all you can give me is a "best practices" response, that will be incredibly helpful. I'm relatively new to the PHP API, but i'm a PHP expert and have been using Filemaker for 20 years.

That being said, my issue is the following:

I have several large databases with 20,000 plus records. All of them respond through the PHP API in less than five seconds when i do a search, or create a record - except for one. It takes 2 to 3 minutes to do a basic insert through a layout that contains only 3 of the fields in the database. We've been pouring over the permissions, logs, etc.

So my questions are:

1. are there any logs that allow me to see exactly what's going on in filemaker as it's happening?

2. are there any best practices docs you can point me to for the PHP API to Filemaker

3. Any common experiences you've had that would cause ONE of the 20 or so databases to be significantly slower than the rest?

Thanks a ton!

Cheers,

Chris

A quick look at you code would be helpful to see if anything problematic sticks out.

I also have a couple of questions. By 'databases' do you really mean 'tables', or do you see the problem when attempting to access any table in a single file? (in older versions of Filemaker each individual file could hold only one table so some developers still refer to tables as databases).

If one file is giving you problems regardless of which layout or table you use the file may be corrupted and need to be repaired. The layout may also have issues. I've seen Filemaker do some goofy things before where simply rebuilding the layout from scratch fixed everything. It's rare but it does happen.

To answer your questions:

1. not really, there's the access log (not helpful here) and the Event log but it doesn't report every transaction, just errors (which might be worth checking), warnings, and "Information" (mostly whether or not backups and scheduled scripts ran )

2. & 3. I think there's a white paper on Filemaker's site Jonathan Stark has some useful information on his site, and some of the early posts in this Forum have some good data. I keep meaning to write something up, but I have a day job. That being said the thing to keep in mind is that the API does not create a connection to the Filemaker databases. So the returned Filemaker object is not a resource handle. The API is simply a series of objects which interface with Filemaker's XML feed via cURL requests. So the limiting step is generally the amount of time it takes the web server on the Filemaker server to generate the XML page. i.e. Any large data set will slow things down.

HTH

Edited March 27, 200916 yr by Guest