2017 has started with a nearly perfect Q1. We've continually worked to stabilize our systems and report the numbers that display this improvement.
Our systems were largely unaffected by Amazon's February S3 Outage. The only impact to Bonsai was a temporary suspension of automated backups. Clusters continued to serve requests normally, so no downtime was reported for February. This fact was referenced on our status page.
There were two short-lived events in March affecting overall uptime. These were both proximately due to high load on hardware serving primarily free and staging clusters. Our response was to add capacity and migrate clusters to more evenly distribute the load.