upgrading the framework on running cluster.

[samek]

upgrading the framework on running cluster.

[samek]

Anyone ?

[elingg]

Hi @samek, is your goal to simply upgrade the framework scheduler? If yes, it seems that should be possible by updating the app in Marathon. @BenWhitehead can jump in and let us know any additional details.

[samek]

@elingg Yes, just the scheduler.

[kensipe]

@samek Is there anything else to follow up on here?

[samek]

@kensipe I've tried upgrading just the tar and marathon json and it did upgrade.

[keithchambers]

@samek as you discovered, we don't have a story around upgrade at the moment.

Backup and restore was added recently: https://github.com/mesosphere/cassandra-mesos/blob/master/docs/docs/backup-and-restore.md

While admittedly very crude, you should be able to use backup and restore as a method of upgrade once you get the a version of the scheduler that supports it.

[BenWhitehead]

Correct, the version of cassandra didn't change between 0.1.1-SNAPSHOT and 0.2.1-SNAPSHOT, there were some changes to the persisted configuration for the cluster and nodes, but the code should handled updating correctly since it's protobuf.

Things will get much more complicated when the version of cassandra itself is changing.

[triclambert]

This repository is deprecated. Please use the latest supported framework here: https://github.com/mesosphere/dcos-cassandra-service, and provide feedback for us here: https://dcosjira.atlassian.net/browse/CASSANDRA.