4 February 2026 service outage
On Wednesday morning, 4 February 2026, at around 08:30, alerts started appearing on Gosta’s monitoring channels. In addition, several customers reported problems with creating entries, and Gosta was unable to connect to the backend systems.
At around 09:00, it was discovered that Gosta’s entire backend server had crashed due to running out of memory. Shortly after this, an error message was displayed at the top of the Gosta page, and the creation of new visits was disabled. The Gosta team investigated the issue intensively throughout the morning, and at around 11:30 the problem was resolved and the services were restored.
More technical explanation
The observed issues were caused by a combination of several factors:
Wednesday morning was a busy day for documentation, and Gosta was under heavy use
A new feature had been deployed to Gosta on Monday, which in error situations tries to retry speech recognition more quickly so that a draft note would be created faster
The speech recognition service used by Gosta experienced its own outages on Wednesday morning, independent of Gosta
Gosta’s servers were configured suboptimally in terms of scaling and memory usage
As a result of these factors combined, the Gosta user interface generated more and more speech recognition requests. Due to the suboptimal memory usage, this caused the backend servers to crash completely, and after restarting they immediately crashed again, as the retry logic kept generating even more requests.
How we will do better going forward
To prevent this from happening again, we will redesign the retry logic. In addition, server scaling and memory usage have already been reconsidered and improved so that they can better handle Gosta’s growing usage.
The Gosta team apologizes for the service outage on the morning of 4 February 2026, and we are doing everything we can to ensure that a similar incident does not happen again.