On July 29, 2019, at 3:50 PM CDT (UTC-5) we were alerted of increased errors with our search cluster. After investigation our engineering teams found that increased memory consumption on a search cluster caused search queries to slow down or fail for a subset of customers. At 5:59 PM CDT (UTC-5) all search functionality was restored to all customers affected.
Increased memory consumption on a search cluster caused search queries to slow down or fail for a subset of customers.
Our engineers are deploying a fix to mitigate this issue moving forward and we’re refactoring our current search cluster configuration to improve efficiency and reliability.