I did the deed Thursday at 11:46am EST time.
I’d already entered a ticket concerning the breaking of LDAP authentication, so when I no longer could trouble-shoot that problem because of my own stupidity, I added that to the original ticket.
Looking back on it, I think that was a mistake too.
Most of us know Blackboard support needs a little more time coming up to speed on the ins and outs of the Vista architecture. My TSM was confused about what I was reporting, thinking I was suggesting a cause for my LDAP troubles. His emails continued to focus on the networking required for connectivity to the LDAP server. I kept telling him that I had proven connectivity to the LDAP server because it only broke after I added the commercial certs.
Now I had to go through, at length, exactly what my act of stupidity had been and why it was catastrophic. By the 2nd or 3rd email/phone call on Friday morning, he got it.
And rose to the challenge.
Unfortunately this was about the same time I was giving up on it and reconciling myself to rebuilding the entire cluster. Not so horrible considering it was a preproduction system I’d just built.
But, I hated to discourage the Bb team’s enthusiasm. So…
We did an sql update on the sourceid.source directly in the database. Changed it from WebCT to WebCT1. Unfortunately no auth rule exists in the database for which authentication source, WebCT’s global database or my configured LDAP, should be used.
They came up with this idea of using a siapi import to change the mapping globally, because they couldn’t find the exact table in the database. (It’s a long post, but hang with me…)
Well, that had to fail because I hadn’t configured either the luminis or the ims adapter yet, although I did have the glcid…
It’s now Friday afternoon and I am again mentioning how I can just blow away the database and start again. A phone message from my TSM now recommends the very same.
Sigh. Next time they’ll get there faster.
As an aside, those phone call waits were extremely painful. I never got right in to Support… that’s because the rest of us were calling for clarification about the Support Bulletin published on March 8th on the Daylight Savings Drama.
I had to get on that boat too since it looked like our Oracle and Solaris remediation wasn’t enough. After a couple more hours of mapping out a strategy, we went to find out the number of the hotfix containing the JVM fix… only to discover it was 62570 and we’d already installed it in January!!!