Bird et al,
There are two separate issues:
1) The main sever underwent some routine upgrade maintainance around 6 p.m. Since we did not know if anyone would be affected, we didn't even mention it, although we knew there was a slight possibility that someone (like Bird /forums/images/graemlins/wink.gif) might experience very brief outage. Sorry.
2) We had the issue of "scrambled" messages such as Harv noted above and a few others (including me) observed over the past few days. This is a recurring issue, although various steps were taken over the past couple of months to reduce the frequency, severity and duration of these incidents.
Basically, the cause of this "not accessible" or "too many connections" is due to user "spikes" when just too many people are doing too many things at once.
TBN is now getting up to 150 simultaneous users, and more than 500 posts per day. This usually is okay and should be handled by our present capacity.
The problem arises when a FEW of those large numbers of users access extensive, complex searches (whole site) and it takes more than 15 or 20 seconds for the results to come back. This can happen with slow/low memory machines and older operating systems, etc.
This in effect ties up much more than normal bandwidth, and when several do such searches at the same time, the result may be the error you saw. As the searches are completed, or the number of users declines, things return to "normal."
For this second problem the absolute solution would be to put on more bandwidth and server memory. It would almost double our monthly server costs to do so. Since it happens infrequently and only to a few people (only moderators are even supposed to see these error messages), we have elected, so far, to "live with it." We are however looking into various longer term strategies, and will implement them as and if the situation warrants further upgrades.
It really is helpful for you to report such issues here whenever you do experience them.
