This incident has been resolved.

We're monitoring an issue where a traffic spikes caused an increase in GET /subscriber response time as well as increased timeouts. Everything appears to be stable now but we'll continue to monitor the situation.