We experienced a period of downtime across our services this morning and would like to apologize for any inconvenience caused by this. Thankfully, we were able to quickly identify the issue and resolve it, allowing for a full resumption of our services. We’ve compiled a post-mortem below to give you full transparency into what happened.
Note: All times referenced below are Coordinated Universal Time (UTC).
As part of our effort to constantly increase the security of our service, especially when it comes to areas that store customer data, we were introducing individual authorization levels to each product. The result of this would limit the access rights of each product to only the necessary customer information they require from our databases and in turn limit the amount of damage caused if a service was to be compromised.
Due to human error, incorrect credentials were used when deploying these changes, rendering most services unable to access the necessary databases. This in turn caused an outage across our whole platform.
The issue has now been resolved, and as a result, our service is now fully restored and all user data even more secure. However, we take these issues extremely seriously and want to apologize for any negative effect this might have had our your reliance of our services. Moving forward, we will endeavour, as always, to avoid errors of this nature by ensuring all work and credentials are double-checked before deployment.