Apache Error: winnt_accept: Asynchronous AcceptEx failed

This is a problem that was occurring a couple months back.  Our networking software, WhatUpPro is reporting that our moodle server is going down in the early am hrs, and then restarting itself. This happens in the early morning hrs only.

The software reports: 

HTTP is Down at least 5 min on Server: servername:(IP address)
is Up on Server: servername:(IP address) - after between 5 - 20 minutes.

Upon examination of the Apache error log, I noticed a semi re-occurring error. I googled the error and found  this post and this post about why this could be happening. The resolution is to disable the use of AcceptEx by adding the Win32DisableAcceptEx directive to the Apache config file.

Import stuff

• error: [Tue Mar 06 15:50:11 2012] [warn] (OS 64)The specified network name is no longer available.  : winnt_accept: Asynchronous AcceptEx failed.
• error log location: c:\program files\Apache Software Foundation\Apache2.2\logs\error.log
• apache httpd.conf file location: c:\program files\Apache Software Foundation\Apache2.2\conf
• Directive added to httpd.conf.
• Don't forget to restart Apache service after the update to the httpd.conf file.

The server has still been going up and down once a night at the same time.  Looking at this post again, I noticed two additional lines that I had not commented out *activated.  By removing the comment from the start of the Apache line, it activated the line, which in this case turns OFF each of these directives.

Two additional lines (directives) activated

EnableMMAP off
EnableSendfile off

Saved the changes, restarted Apache.  Monitor....

The server is still recycling at the same time each am.  It is the same time each evening, and only once each day, which suggests very loudly that something else, kicking off in the environment at a fixed time, is causing the conflict.  I looked again in the Apache error log but am only seeing the message mentioned above.

What could it be, what is running on this server at a fixed time? Could something in moodle be trying to run at a fixed time?  There are scheduled tasks, which run every 30 minutes for each of the 10 moodle instances on this server (vhost).  These are a very normal part of moodle and need to stay scheduled.  Is it a virus scan on the server?  The Lakenet guy said the virus scan runs at a different time each night.  I looked again at other services running on this server and found a mysql service that was set to automatic sitting around.  This service was left sitting around after I uninstalled mySQL client and server from the server.

I removed the mysql windows service using this post as a guide.