Ask HN: Slow OpenAI Inference on AWS Bedrock

Inference on AWS Bedrock for OpenAI models has been very slow and glitchy the last few days with today being the worst. Very long latency of over 5 minutes to first token. What's going on?