A bad case of the Mondays
On Monday, the server that hosted our main web site, including this blog and our user forum, had a major catastrophe and died. We spent the rest of the day getting a new server up and running to take its place. Perhaps it was a sort of comeuppance as we were recently congratulating ourselves on having launched Dabble DB a year ago and having had no system-wide outages — scheduled or otherwise — in that entire time.
Unfortunately, the server that went down was also responsible for directing requests to the various servers that host customer databases, so everyone lost access to their databases for a time too. No public web site, no blog, no forum, no database access... all in all, not our finest hour.
Naturally there are lessons to be learned after such disasters. We've identified some of the weaknesses in our old system and are taking steps to make sure that when the next server fails, it will be a less calamitous affair. Our systems will be more independent so that we can better communicate with you in the event of emergency. This blog, for instance, is now hosted off-site by the fine folks at TypePad, so it should always be available for us to give status updates. (You may want to change your bookmarks to http://blog.dabbledb.com/).
I'm proud to say that in all the mayhem, we lost absolutely no data. A little hair and some sanity on the other hand... aah, who needs 'em?