[IGNITE-9738] Client node can suddenly fail on start

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.8
• Assignee: Vladislav Pyatkov
• Reporter: Vladislav Pyatkov
• Created: 28/Sep/18 14:58
• Updated: 22/Oct/18 19:20
• Resolved: 22/Oct/18 15:20

Description

If client joining to large topology it can to spend some time on waiting TcpDiscoveryNodeAddFinishedMessage, but in that time it can not to send TcpDiscoveryClientMetricsUpdateMessage. By that reason server can to reset client from topology.

Client node considered as unreachable and will be dropped from cluster, because no metrics update messages received in interval: TcpDiscoverySpi.clientFailureDetectionTimeout() ms. It may be caused by network problems or long GC pause on client node, try to increase this parameter. [nodeId=a3493895-7c13-403c-bf0d-94eab0000011, clientFailureDetectionTimeout=30000];

We should to sent TcpDiscoveryClientMetricsUpdateMessage as soon as possible without, waiting finish of join procedure.

Issue Links

• links to: GitHub Pull Request #4968
• links to: MTCGA