Writes to tables are delayed

Our systems have fully recovered and all table updates have been processed.

We have confirmed that there is no data loss on writes, simply delays on persisting the writes through to tables.

We have confirmed that our autoscaling infrastructure is doing its job and working through the table write backlog, but did not respond fast enough in this instance.

Our systems are recovering and we are continuing to monitor.

We have identified a problem where our infrastructure that supports table writes has fallen behind processing the writes, leading to a backlog of table writes.

We are reviewing our system telemetry to attempt to identify where the write contention bottleneck is in our infrastructure.

We are receiving reports that saving data to tables is currently delayed. We are investigating.