Grigory Domozhirov created IGNITE-10317:
-------------------------------------------
Summary: “Failed to do service reassignment” on server stop with singleton service
Key: IGNITE-10317
URL:
https://issues.apache.org/jira/browse/IGNITE-10317 Project: Ignite
Issue Type: Bug
Affects Versions: 2.5
Reporter: Grigory Domozhirov
I have Ignite cluster with 2 server nodes and singleton service deployed on it.
If I first stop node which currently runs service, and then (with particular delay, about ~ 100 ms) second one, I'm getting
{code:java}
ERROR o.a.i.i.p.s.GridServiceProcessor - Failed to do service reassignment (will retry): statService
org.apache.ignite.internal.IgniteInterruptedCheckedException: Got interrupted while waiting for future to complete.
at org.apache.ignite.internal.util.future.GridFutureAdapter.get0(GridFutureAdapter.java:185)
at org.apache.ignite.internal.util.future.GridFutureAdapter.get(GridFutureAdapter.java:140)
at org.apache.ignite.internal.processors.service.GridServiceProcessor$TopologyListener$1.run0(GridServiceProcessor.java:1810)
at org.apache.ignite.internal.processors.service.GridServiceProcessor$DepRunnable.run(GridServiceProcessor.java:2027)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
{code}
I guess, second server sees another server is down, so it wants to start service. Then at unlucky moment this server stops, and {{ExecutorService depExe}} at {{GridServiceProcessor}} get interrupted, so exception occurs. But Ignite knows if it stopping, so I'm sure Ignite could expect interrupts, or take care of it in some way.
Ignite version: 2.5.2
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
Locked
