Quick incident summary for this morning's processing delay:

• Delay began at approximately 8:00:00am PDT

• The delay peaked at 141 seconds at approx 8:06:50am PDT

• The delay then slowly dropped and cleared at approx 8:20:40am PDT

The immediate cause appears to be high load on the database during this time. In the short term, we're investigating how to move this workload so that it doesn't affect the processing pipeline.

Processing delay has returned to normal levels (approx 1 second).

Processing delay is at 47 seconds; taking a bit longer than expected to return to normal levels but still continuing to clear. We're watching the system closely to understand the bottleneck.