You may remember several weeks ago that Calliflower experienced two service outages.  At the time we diverted all the resources that we had to fixing those outages.  Today I’m writing to let you know the progress we’ve made.

Firstly, we discovered some remarkable things about the code we’ve written.  It’s resilient, tough, and very hard to bring down.  The downside is that inefficiencies we had introduced into Calliflower through a couple of recent updates had resulted in very long response times for some classes of activities.  For example, during debug we discovered one scenario where Calliflower might take several minutes to respond to certain kinds of user requests.  At no time did it actually cease functioning, it simply got slower and slower!

The team has been hard at work over the last few weeks optimizing that code, and the optimizations were deployed last week.  I’m happy to report that no request should take longer than one second now, and that we’ve made great strides in improving the overall responsiveness of Calliflower.  Most importantly, we won’t see a repeat of what happened three weeks ago.

As one person remarked to me, these growing pains are signs that the word is getting out, and that we’re succeeding as a business. True! Nevertheless, everyone here has had a renewed focus on customers as a result of these incidents.

Thank you to all of our customers for the patience you’ve shown.  We’re grateful for your continued patronage.

Alec Saunders, CEO