6GB error log – mod_fcgid: stderr

The PHP warning is the probable cause of the serialized strings in the Apache error log. The warning could be caused by many things, but the most likely is your exceeding some type of resource limit.

You should contact your hosting company. In addition to fixing the errors, they should separate the PHP error log from the Apache error log — it’s a simple change in the PHP configuration.

Hi, thanks for your reply.

so you don’t think it’s a specific plugin, more server related?

The website is hosted on an “unmanaged” VPS which I have full access to although as it’s unmanaged their support only covers certain things.

the server is fairly high spec so it would seem strange a WP site would be using to much – 4 cores / 6GB RAM / 100GB SSD.

I have also registered with Loggly but as the file is 6GB it’s to big to upload.

I will contact them anyway as the log is now 11GB !

thanks

You may well have enough server power, but an incorrect Apache/MySQL/PHP configuration can create resource limit issues.

Without seeing your server logs, the best I can provide is an educated guess: your MySQL configuration is causing problems for PHP. For example, insure the max_allowed_packet setting in my.cnf is large enough to handle whatever WooCommerce is throwing at it. (This is just one of many possible settings to inspect.)

You should also set error_log in your php.ini file. PHP will then write its logs to the file of your choice and not to the Apache error log.

And until you get this resolved, I strongly suggest that you set a CRON task to truncate the error log hourly. A log file that large will cause server issues.

thanks for that, I will investigate all the things you suggest.

I have set the LOG rotation to daily – as originally it went to 60GB and crashed my server which in turn corrupted the WP database, so I had to re-install an older version of the site.

One of the log files that came in this morning has a ‘legible’ section at the end… do you know what below means? I’m running oceanWP, do you think the theme or core file is corrupted?
thanks.

‘ WHERE option_name = ‘_transient_wc_layered_nav_counts_pa_disk-type’ made by require(‘wp-blog-
header.php’), require_once(‘wp-includes/template-loader.php’),
include(‘/plugins/woocommerce/templates/archive-product.php’),
do_action(‘woocommerce_before_main_content’), WP_Hook->do_action, WP_Hook->apply_filters,
OceanWP_WooCommerce_Config->content_wrapper, get_template_part, locate_template, load_template,
require(‘/themes/oceanwp/woocommerce/wc-content-wrapper.php’), do_action(‘ocean_before_primary’),
WP_Hook->do_action, WP_Hook->apply_filters, oceanwp_display_sidebar, get_sidebar, locate_template,
load_template, require_once(‘/themes/oceanwp/sidebar.php’), dynamic_sidebar, WP_Widget->display_callback,
WC_Widget_Layered_Nav->widget, WC_Widget_Layered_Nav->layered_nav_list, WC_Widget_Layered_Nav-
>get_filtered_term_product_counts, set_transient, wp_cache_set, W3TC\\ObjectCache_WpObjectCache->set,
W3TC\\ObjectCache_WpObjectCache_Regular->set, W3TC\\ObjectCache_WpObjectCache_Regular-
>_transient_fallback_set, update_option, W3TC\\DbCache_WpdbNew->update,
W3TC\\DbCache_WpdbInjection_QueryCaching->update, W3TC\\_CallUnderlying->update,
W3TC\\DbCache_WpdbNew->update, W3TC\\DbCache_WpdbInjection->update, W3TC\\DbCache_WpdbNew-
>default_update, W3TC\\DbCache_WpdbNew->query, W3TC\\DbCache_WpdbInjection->query,
W3TC\\DbCache_WpdbNew->default_query

Actually I think that W3TC is the culprit. Does the problem go away, even temporarily, if you completely purge its cache? If so, give some thought to increasing its cache size, or perhaps finding a new/different cache plugin.

If the problem persists, and your site ran fine immediately after purging the W3TC cache, then you might want to consider temporarily deactivating W3TC. Make sure to first check the W3TC documentation to insure it doesn’t delete configuration settings when it is deactivated!

Hi, just so you know, I removed and deleted W3TC cache. The problem still persisted – massive 10GB log file every day.

So the next day I removed the ‘WooCommerce Admin’ plugin and now the logs have gone back down to a few MB a day.

So I am thinking it was the woocommerce admin plugin causing the conflict, a bit strange as it has been installed for many months with no issue.

I won’t be using this plug anymore.

anyway thanks for your input.