All applications hosted on our main cluster, as well as the Bubble editor, experienced slower load times as well as load failures during this incident. The cause of the incident was a failure of one of our redis caching cluster shards, and we resolved it by failing over to the backup server we have on standby. The root cause of the failure was an increase of load on the shard caused by some bad logic in our code for fetching application data. Stabilizing our redis cluster is an ongoing operation, and redis cluster issues were also responsible for the incident on Jan 9th. Since that incident, we have put in place some temporary mitigations, but are continuing to work on making sure our overall redis cluster is sufficiently protected from load spikes and appropriately provisioned.