We're currently verifying the status of our services after the switch to the encrypted cluster.
Posted Apr 21, 2018 - 14:05 CEST
In progress
Scheduled maintenance is currently in progress. We will provide updates as necessary.
Posted Apr 21, 2018 - 13:01 CEST
Scheduled
What should be done?
The master database of the global cluster needs to be encrypted. This improvement of the database relates to the new EU legislation ISO27001 / GDPR. We've already prepared this change. We're running an encrypted shadow cluster that's replicating from the current database. During the maintenance window we need to stop replicating and then reroute traffic to the shadow cluster. Because we've prepated it the entire thing should take us only 10 minutes, but we want to make sure it all goes well and create a backup snapshot it takes a bit longer, but there's no impact from that. We've already tested this change on our staging environment. Besides this we also upgraded our new cluster so we can improve on performance for future requirements.
What is the possible impact?
During the rerouting we need to put the database in read-only mode. We don't want to miss any data. RO-mode should last up to 5 minutes maximum. After that everything will be back to normal operating mode. In case something goes wrong, we can switch back to the old situation within 30 seconds. (maximum TTL for DNS changes)