We've mitigated the secondary impacts and are deployed in an additional data center while we root cause the latency spikes.

One of our datacenters experienced nine 2-3 minute periods of high latency which sometimes caused additional latency and call creation failures for 5-7 minutes during recovery.