It turns out that my instinct concerning our deployment date was correct.  They spent most of yesterday getting that last component running.  They had missed some more files when they released the code to the read-only filesystem.  They had also misconfigured some settings that took a while to track down.  This gets pretty frustrating, because the instructions for those settings are clearly given in the installation documentation.  The other thing that is frustrating is that the deployment people aren’t giving our application their full attention.  The original person got called off to work a severity 1 problem for someone else, so they brought in someone new.  While he was competent and did manage to resolve the problem, it took a long time because he wasn’t dedicated to our project.  He was constantly being interrupted by other issues.

The final result yesterday was that we found some further misconfigurations in the settings that will require a restart of the servers to fix.  Because of the rules that they impose, restarts have to be done at night and they have to get permission ahead of time.  This means that we will only be able to restart the servers tonight.  Therefore, I have to be online at 9:00pm tonight.  I can only hope that they work faster than they did the night before.  Provided they have fixed the settings correctly, I will be able to get on the new system and run some testcases to verify that it is functioning correctly.  Once that is done, we’ll let the customer know that it’s ready and they can start their acceptance testing.  If that works out, the proxy changes will be done next Thursday night to make the site live. 

Of course, given the problems we’ve had so far, I’m not even going to hazard a guess whether we’ll actually make it this time (the original schedule called for user acceptance testing to complete tomorrow morning with go-live tomorrow afternoon).

This means that I may have to cut my shooting session short today, in order to make it home in time to get online.