Today I applied the infrastructural updates and they failed (postsetupconfigurationtaskexception in step 8 of 9). There were several messages in the diagnostics log (B2b_Upgrade) and upgrade log (path not found). In our case the upgrade log complained about a path not found for a specific site. If you would browse to that site, you will receive a http 400 bad request error. If you try to find the site using stsadm, it will fail (backup, delete etc). So how to get rid of that site? The answer is simple, through Central Admin –> Application management –> Manage Content Databases. Now select the appropriate web application from the drop down list, next remove all content databases (don’t worry, databases will not be deleted). Make sure that you have written down which content databases were attached and where (e.g. databaseserver name or clusternode\instancename). Next, add the content databases again using the same page in Central Admin. This will clean up things for you. Now try to run the Configuration wizard again on each front end.
Someone else had a similar problem with a different cause. In their case the web.config was missing from the layouts directory. Try copy the web.config file from another web servers layouts directory or use the one in the config directory. Another person said the web.config file contained question marks and he solved it by fixing those.
As you can see, same message, different causes and different solutions. Hope this one may be of help to you.