I have written a post some time ago about the LLAWP process and how it could ruin your happy engineer life. There are new news from a front. So again, you have IIS service, Siteminder installed and configured, let’s imagine you need to reconfigure your Siteminder agent to another policy server. You use to use for these purposes smwebagent.conf file inside of ~/config/IIS folder in Siteminder’s installation file structure. You have changed whatever you had to change. Open a browser, and goto to a Siteminder-enabled website. And you see the 500 error. What to do? Your common sense says “Go, restart IIS”. You have done so. Again, the browser, refresh the site. You see a login box from your new login server. Awesome! Let’s login with a for-sure-existing credentials. And your login server says “Internal error”. What is that? Go lookup server’s event log – nothing! Fallback on Siteminder’s log file… If it was disabled you have to enable it and repeat the login ceremony again. Bother! Then inspect the logs, where Siteminder dumps its configuration. What a hell! It shows you an old policy servers and login servers urls! Ok, let’s suppose something is cached. So something has to be killed to free the in-memory cache. What would that be? Who is the suspect? Oh, I remember that buggy process called LLAWP. Go and shut him down.
You come back to your browser – and voila! A nice face of the site is smiling to you