Over the past month or two we have concentrated almost all our efforts on improving the stability and reliability of Treejack and Chalkmark. I’m pleased to say that I think we’ve made great headway so far! Of course this wasn’t an initiative pulled out of the air.. we were having some significant growing pains. We hit technical limitations on all fronts, almost all at once. Our database server ran out of space, our automated backups were taking too long, we found some things in the database that really needed indexing, we were critically low on memory at busy times, all your Chalkmark images were becoming a big storage issue, our monitoring services weren’t behaving as expected… the list does go on and gets increasingly technical.
So we’ve taken action on all these fronts. Servers have been upgraded, code has been made more efficient and monitoring has been improved. There is now very little chance that I can possibly sleep through a server crash (awesome!), but fortunately there’s substantially fewer ways that could happen (awesome twice!). I hope too that you’ll be noticing a substantial performance improvement; things should generally be snappier for both survey administrators and participants in both Treejack and Chalkmark. Please let me know how you’re getting on- we’re a small team here and we take your feedback straight to heart. You will be heard and your comments are extremely useful to us.
Thanks to everyone who let me know when things weren’t quite right over the past couple of months. We’ve maintained a great percentage of uptime, but even 99.5% uptime allows for 43.8 hours downtime, to which we haven’t even come close, but it has made me think about people who boast their 99% uptime…!