Dear all

I am using search files since a few months ago and it is perfect for me due to the fact that it searches for information in files located in a the "files" folder rather than searching for files linked to specific nodes.

However, I need to search in many files (right now I have uploaded mode that 10.000) and, since a few weeks ago, the system has started to fail when indexing.

These are some of the details about my database:

• search_dataset: 4610 rows (49,4 Mb table size)
• search_files_files: 10.419 rows (2,2 Mb)
• search_index: 1.652.632 rows (133,5 Mb)
• search_total: 151.193 rows (6,2 Mb)

• search_directories: 1 row
• search_directories_files: 0 row
• search_files_helpers: 4 rows
• search_node_links: 12 rows

When I try to access "search settings" in the admin panel, the system gets very slow ant it returns the following errors:

Fatal error: Maximum execution time of 60 seconds exceeded in D:\xampp\htdocs\observatorios\includes\database.mysqli.inc on line 108

Fatal error: Maximum execution time of 60 seconds exceeded in D:\xampp\htdocs\observatorios\includes\session.inc on line 56

Fatal error: Maximum execution time of 60 seconds exceeded in D:\xampp\htdocs\observatorios\includes\session.inc on line 14

Sometimes we upload many files during the day. My application is installed in an intranet and I run the cron only at night to avoid delays for the user.

I call the cron.php every half an hour from 0:00 to 7:00 in order to make sure new contents are indexed. I get the following log message: Cron has been running for more than an hour and is most likely stuck.

My question is:

* Is my application too demanding for search_files + mysql? Should I try a different module or a different approach to index files? or is there any configuration issue that neds to be taken into account?

Regards

Comments

geme4472’s picture

geme4472 commented

I have been struggling with this issue with this module, search_attachments, and swish-e. In this particular case, once there's a timeout, cron will not rerun the following time because you're "Attempting to re-run cron while it is already running." The only issue is that your original cron run timed out. If you delete the cron_semaphore variable in the variables table, you'll be able to run cron again. Search_files will also abide by your #nodes to index setting in search settings, so you can ratchet that down. Not that that's a solution... I'll be working on this, but if you come up with anything (perhaps a register_shutdown_function?), please post.