Monday turned out to be a terrible day for me.  Not only did the server not stand up to the test of a WWDC live-blog, but it was 100% out of my hands.  This past Friday, I knew I had to upgrade our server to be able to handle the huge traffic spike we were expecting the following Monday (yesterday) on Appletell.  I didn’t want to take any chances, so I had them up the amount of RAM by four times.  Since they had to take the server offline so they could put it on their workbench and insert the sticks of RAM, I sent out a note to the D-team letting them know we would be down sometime between 2:00AM and 7:00AM Eastern on Saturday morning.  As far as I knew, all of the work got completed as scheduled.

Then comes Monday morning.  The keynote started at 10:00AM Pacific, and traffic started picking up at around 9:00AM—enough to be considered a traffic spike.  All seemed to be going well, until about 10:05AM when our sites started to hang—blank white page and all.  I called up our web host and asked them if they could do anything to help.  I told them how I can’t believe this is really happening since I just increased the amount of CPU by four times!  The technical support rep. said…“What do you mean?  I don’t see any upgrade”.  They never…installed…the RAM… which was something I couldn’t believe I was hearing.  Furiously, I said “let me talk to your supervisor”.  He then IM’d for a minute, and said…“we can install the RAM right now—it should only take about 15-20 minutes.  Great.  Not only were we down in the middle of a HUGE traffic period, but we had to physically take our server offline too.

At about 10:45AM when we got the server back up, it seems the thousands of people that were with us had already found another source for their Steve Jobs keynote coverage, so we were only getting late-comers.  Understand now why Monday was a terrible day for me?