Introduction
On January 23rd, between 11:05 and 12:25 CET, the Tink platform experienced an outage linked to issues reported on December 19th.
Root Cause Analysis
At 10:29 CET, Tink proceeded with a planned maintenance operation to strengthen the reliability of its distributed database infrastructure. Operations were successfully executed on multiple nodes but failed on one node, which entered into an error state directly impacting the overall database availability.
Remediation
At 12.20 CET the faulty node was temporarily excluded from the database cluster, and impacted products became immediately available again. The faulty database node was later on permanently removed in a controlled manner during a planned maintenance window.
Follow-up actions
In addition to the preventive actions identified earlier in December, Tink will ensure that such operations shall be conducted under planned maintenance windows to minimise customer impact.