Mews experienced a notice incident on April 20, 2026, lasting 3h 17m. The incident has been resolved; the full update timeline is below.
Update timeline
- investigating Apr 20, 2026, 09:25 AM UTC
We have noticed a downgraded performance of the system. We are currently looking into it. Thank you for your patience.
- identified Apr 20, 2026, 09:43 AM UTC
We believe we have narrowed down the cause and remediated this. We're observing an improvement in performance and are continuing to monitor this closely.
- resolved Apr 20, 2026, 12:42 PM UTC
The issue is now fully resolved. Mews is now operating as expected.
- postmortem May 11, 2026, 07:02 AM UTC
## Problem On **20 April 2026**, a newly enabled pricing-related flow caused a database performance issue that resulted in a **full production outage**. Between **09:03 and 09:30 UTC**, customers were unable to use Mews. During that period, users experienced login failures, guests could not view room prices or complete bookings through the Booking Engine, and some distribution partners may have missed reservations if their systems did not support retries. We’re sorry for the disruption this caused. ## Action ### Immediate response After identifying the issue, we disabled the related feature flags and scaled up the production database to restore performance. ### Completion The customer-facing outage ended at **09:30 UTC**, and we continued monitoring the system until the incident was fully resolved at **12:49 UTC**. ## Causes ### What went wrong The newly enabled flow triggered database queries that did not perform as expected under production traffic. This caused a rapid increase in database load and led to the outage. ### Why we didn’t catch it earlier The issue did not appear in lower environments, where traffic levels were not high enough to expose the problem before production release. ## Solutions We are implementing the following improvements to reduce the likelihood and impact of similar incidents in the future. ### 1. Stronger database and query review We will strengthen how we review database queries and schema changes during development so performance risks are identified earlier. ### 2. Safer production rollouts For higher-risk changes, we will use a more gradual rollout approach and closer monitoring during initial production enablement. ### 3. Better test coverage in realistic environments We will improve performance testing so new feature paths are exercised in conditions that better reflect production usage. We take the reliability of Mews seriously and are strengthening our rollout, testing, and performance review practices to reduce the chance of this happening again.