Vladimir Ozerov created IGNITE-7472: ---------------------------------------
Summary: Exchange initiated by custom disco events could hang in case of coordinator leave Key: IGNITE-7472 URL: https://issues.apache.org/jira/browse/IGNITE-7472 Project: Ignite Issue Type: Bug Components: cache Affects Versions: 2.4 Reporter: Vladimir Ozerov Fix For: 2.5 Reproducer: {{WalModeChangeAdvancedSelfTest.testServerRestartCoordinator}} WAL mode change operation initiates an exchange through custom discovery event. The test does the following: 1) Initiate constant flow of such events from one node 2) Constantly stops current coordinator node and start another one sever node Debug shows the following: 1) Remaining client nodes is on exchange X, waiting for new coordinator to send affinity message 2) New coordinator is on exchange [X+1], waiting for client to send partition message -- This message was sent by Atlassian JIRA (v7.6.3#76005)