Repairs get stuck with IN_QUEUE/ON_TIME status #1189
Description
Detailed description:
• What is happening?
When deploying the new ecChronos agent configured as host aware, the on demand repair gets stuck with IN_QUEUE status and 0% repaired.
Command used: ecctool run-repair --keyspace ks --table tb2 --all
The last log messages are these
[http-nio-127.0.0.1-8081-exec-1] [org.apache.catalina.core.ContainerBase.[Tomcat].[localhost].[/]] Initializing Spring DispatcherServlet 'dispatcherServlet' [http-nio-127.0.0.1-8081-exec-1] [org.springframework.web.servlet.DispatcherServlet] Initializing Servlet 'dispatcherServlet' [http-nio-127.0.0.1-8081-exec-1] [org.springframework.web.servlet.DispatcherServlet] Completed initialization in 1 ms [http-nio-127.0.0.1-8081-exec-3] [com.ericsson.bss.cassandra.ecchronos.core.impl.repair.scheduler.ScheduledJobQueue] Adding job: Vnode On Demand Repair job of ks.tb2, Priority: HIGHEST
These are some configurations I have in ecc.yaml (besides the default values):
connection:
cql:
agent:
type: hostAware
localDatacenter: datacenter1
contactPoints:
- host: localhost
port: 9042
hostAware:
hosts:
- host: localhost
port: 9042
• What did you expect to happen?
On demand repair to complete.
• What have you tried?
Retrigger the repair, and trigger a repair in a different agent.
• What version of ecChronos are you using?
Build from agent/master branch.
• Was the problem detected during an upgrade or downgrade procedure?
No.