Yesterday, we experienced about 1 hour of intermittent downtime of the bots of Chatlayer.ai in production. The instability was caused by a massive unexpected surge in requests, most of which were infinite loops caused by Chatlayer bot to external bot interactions.

To mitigate the problem, we blocked the problematic sessions and upscaled the resources in our cluster, which resolved the problem.

To avoid this problem in the future we will:

• Implement automatic detection of infinite loop conversations & the blocking of these
• Allow for more flexible scaling of our cluster to better support surges in load.