Alexey Goncharuk created IGNITE-5560: ----------------------------------------
Summary: A failed service must be redeployed when possible Key: IGNITE-5560 URL: https://issues.apache.org/jira/browse/IGNITE-5560 Project: Ignite Issue Type: Improvement Components: managed services Affects Versions: 1.7 Reporter: Alexey Goncharuk Fix For: 2.1 I observed the following behavior: when a service deployment (or run) fails with an unexpected exception, Ignite just outputs a warning to the console and does not attempt anything to recover. In our deployment, we rely on a cluster singleton to be always present in a cluster. If a service fails, Ignite should attempt to failover this service to some other node, if possible -- This message was sent by Atlassian JIRA (v6.4.14#64029)