View source for Apache log errors: Invalid argument: AH03099: ap queue pop failed

<!-- seo -->
{{#seo:
|title=Apache log errors Invalid argument: AH03099: ap_queue_pop failed
|titlemode=replace
|description=This error may be to do with rewrite rules, not Event MPM settings!
|og:image=https://www.is-beer-a-vegetable.com/wiki/File:Log_errors.png
|og:type=article
}}
<!-- /seo -->
==={{PAGENAME}}===
[[Image:log errors.png|300px|thumb|right|{{PAGENAME}}]]
I recently made some changes to my Apache config and then noticed thousands of similar errors appearing in my Apache error_log.  The time intervals for these errors were virtually identical.

===Examples of repeated errors===
<pre>
[Wed Jan 18 01:28:05.275815 2017] [mpm_event:crit] [pid 4844:tid 140271072855808] (22)Invalid argument: AH03099: ap_queue_pop failed
[Wed Jan 18 01:28:05.275818 2017] [mpm_event:crit] [pid 4844:tid 140271072855808] (22)Invalid argument: AH03099: ap_queue_pop failed
[Wed Jan 18 01:28:05.275821 2017] [mpm_event:crit] [pid 4844:tid 140271072855808] (22)Invalid argument: AH03099: ap_queue_pop failed
</pre>

I had never had these problems before I made the changes.

Googling for the problem showed a few suggestions, but no proven solutions. Most of the suggestions were around Event MPM settings - not surprising as that was where the errors are being issued from.

The changes that seemed to cause the problem were rewrite rules from http to https (ssl) 

The problem rules
<pre>
<VirtualHost *:80>
-    RewriteEngine on
-    ServerName       www.domain-example.co.uk
-    ServerAlias      domain-example.co.uk
-    RedirectMatch permanent ^/(.*) https://www.domain-example.co.uk/$1
 </VirtualHost>
</pre>
<GoogleBanner></GoogleBanner>

Another clue was that non-secure (http) calls to my sites were causing entries in my default server http access_log so doubling the resources used for each page.

'''If you have a similar problem, take a look at any rewrite rules you may have changed!'''

[[Category:Useful information]]