Recently, using Apache as a load balancer for a cluster, according to the relevant configuration, after adjusting the relevant content, the Apache httpd.conf file is adjusted. Add the following content:
<virtualhost *:80>proxyrequests Offproxypass/balancer://lfcluster/stickysession=jsessionid Nofailover=off <proxy balancer://lfcluster>balancermember http://172.18.2.80:6888 loadfactor=1 route=proxyserver1 Balancermember http://172.18.2.63:6888 loadfactor=1 route=proxyserver1 </Proxy></VirtualHost>
After adding such as content, the original normal Apache can not start!
Looking at Apache's backend log error.log, we found the following:
[Tue Dec 09 10:34:51 2014] [Notice] parent:received Restart Signal--restarting the server.
Syntax error on line 499 of c:/apache2.2/conf/httpd.conf:
Balancermember can not has a balancer name when defined in a location
[Tue Dec 09 10:34:51 2014] [Notice] Child 3432:exit event signaled. Child process is ending.
[Tue Dec 09 10:34:51 2014] [Warn] (OS 995) I/O operations have been aborted due to thread exits or application requests. : Winnt_accept:asynchronous AcceptEx failed.
It looks like the problem with these configurations has just been increased, and 499 lines of problems occur, that is:
The exception is: Balancermember can not has a balancer name when the defined in a location. Look carefully, and Google, to confirm that some places may be inaccurate characters caused.
So the content of the line, found the last pair of a space!
Remove space, save, restart Apache.
Ok!
Apache fails to start as a configuration problem for the load balancer