[FLINK-13750] Separate HA services between client-/ and server-side

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Done
• Fix Version/s: 1.10.0
• Component/s: Command Line Client, Runtime / Coordination
• Labels: pull-request-available

Release Note

The `HighAvailabilityServices` have been split up into client side `ClientHighAvailabilityServices` and cluster side `HighAvailabilityServices`. When implementing your own high availability services, you should follow this separation by overriding the factory method `HighAvailabilityServicesFactory.createClientHAServices`. Moreover, `HighAvailabilityServices#getWebMonitorLeaderRetriever` should no longer be implemented since it has been deprecated.

• Assignee: Zili Chen
• Reporter: Chesnay Schepler
• Created: 16/Aug/19 10:06
• Updated: 02/Oct/19 17:47
• Resolved: 04/Sep/19 14:10

Description

Currently, we use the same HighAvailabilityServices on the client and server. However, the client does not need several of the features that the services currently provide (access to the blobstore or checkpoint metadata).

Additionally, due to how these services are setup they also require the client to have access to the blob storage, despite it never actually being used, which can cause issues, like FLINK-13500.

Tison Would be be interested in this issue?

Issue Links

• causes: FLINK-13500 RestClusterClient requires S3 access when HA is configured
• is blocked by: FLINK-13912 Remove ClusterClient#getClusterConnectionInfo
• relates to: FLINK-13960 Provide default implementation for deprecated HighAvailabilityServices.getWebMonitorLeaderRetriever
• supercedes: FLINK-13500 RestClusterClient requires S3 access when HA is configured
• links to: GitHub Pull Request #9509
• links to: GitHub Pull Request #9609