Scheduled - Instaclustr are performing operating system replacements across all clusters over the coming months. While this is a relatively significant change, the impact to our customers is expected to be very similar to our regular operating system patching program. That is, the change will require a rolling restart of each cluster, however this will be completed without loss of quorum for all Cassandra and Kafka clusters and so we advise that customers should not see an interruption to service.

For all PRODUCTION clusters, and clusters running Spark, customers will receive a further notification prior to patching to confirm a specific maintenance window.

Schedules for the maintenance have been sent out individually to all customers, and we advise that customers refer to the scheduling included in this notification for more accurate information on their specific maintenance timeframe.

Further background information:

Instaclustr is in the process of scheduling all of our nodes to undergo a base operating system migration from CoreOS to our latest platform offering, which utilises Debian. We have been planning this migration activity for quite some time and we are now pushing forward with the replacement. This activity has been necessitated by the forthcoming end of life for CoreOS (End-of May 2020) but will bring us advantages such as earlier access to new cloud provider offerings and greater availability of relevant public security benchmarks. We have conducted thorough functionality and performance testing prior to moving to the deployment phase of this change and do not expect any adverse impacts. We are also following an incremental roll-out approach across our managed fleet (as outlined above) and will be closely monitoring for any issues through that roll-out.
Apr 24, 02:39 UTC