Mews incident

Mews POS unavailable

Resolved after 27 minutes, on October 2, 2026.

postmortem Oct 08, 2026, 08:05 AM UTC

**Problem** On 2 October 2026 at 15:14 CEST, POS became slow or unavailable for all customers for approximately 27 minutes. Customers could not log in or complete transactions. **Action** The team identified the cause and reversed a configuration change within 20 minutes of the first alert. Service was fully restored by 15:41 CEST, and operations returned to normal shortly after. **Causes** An unexpected increase in incoming payment notifications put excessive load on the POS database, which slowed down and then interrupted the service. **Solutions** The configuration change was immediately reversed to stop the notification flood. We are now improving the notification handler to filter out irrelevant messages before re-enabling the feature, with a staged rollout plan that includes database monitoring at each step.

Started
Oct 02, 2026, 01:14 PM UTC
Resolved
Oct 02, 2026, 01:41 PM UTC
Duration
27m
Detected by Pingoru
Oct 02, 2026, 01:14 PM UTC

Update timeline

  1. resolved Oct 08, 2026, 08:04 AM UTC

    On 2 October 2026 between 15:14 and 15:41 CEST, Mews POS was slow or unavailable for all customers. Customers could not log in to POS or complete transactions. Our team identified the cause and reversed a configuration change within 20 minutes of the first alert. Service was fully restored at 15:41 CEST, and operations returned to normal shortly after.

  2. postmortem Oct 08, 2026, 08:05 AM UTC

    **Problem** On 2 October 2026 at 15:14 CEST, POS became slow or unavailable for all customers for approximately 27 minutes. Customers could not log in or complete transactions. **Action** The team identified the cause and reversed a configuration change within 20 minutes of the first alert. Service was fully restored by 15:41 CEST, and operations returned to normal shortly after. **Causes** An unexpected increase in incoming payment notifications put excessive load on the POS database, which slowed down and then interrupted the service. **Solutions** The configuration change was immediately reversed to stop the notification flood. We are now improving the notification handler to filter out irrelevant messages before re-enabling the feature, with a staged rollout plan that includes database monitoring at each step.

Your product runs on more than Mews

Add the vendors you depend on and get one message when any of them reports trouble. Free for 3 vendors, no card.

Add your stack →