[jira] [Created] (KNOX-2869) Possible NPE at CM cluster configuration monitor startup

["Sandor Molnar (Jira)" <ji...@apache.org>]

Sandor Molnar created KNOX-2869:
-----------------------------------

             Summary: Possible NPE at CM cluster configuration monitor startup
                 Key: KNOX-2869
                 URL: https://issues.apache.org/jira/browse/KNOX-2869
             Project: Apache Knox
          Issue Type: Bug
    Affects Versions: 2.0.0
            Reporter: Sandor Molnar
            Assignee: Sandor Molnar
             Fix For: 2.1.0


In some very rare cases, it could happen that CM service discovery-related configuration file(s) were wrongly serialized on the file system and the next time Knox starts the service initialization fails like this:
{noformat}
2023-01-30 10:39:01,733 FATAL knox.gateway (GatewayServer.java:main(193)) - Failed to start gateway: java.lang.NullPointerException
java.lang.NullPointerException
        at java.base/java.util.concurrent.ConcurrentHashMap.computeIfAbsent(ConcurrentHashMap.java:1690)
        at org.apache.knox.gateway.topology.discovery.cm.monitor.ClusterConfigurationCache.addDiscoveryConfig(ClusterConfigurationCache.java:73)
        at org.apache.knox.gateway.topology.discovery.cm.monitor.ClouderaManagerClusterConfigurationMonitor.loadDiscoveryConfiguration(ClouderaManagerClusterConfigurationMonitor.java:186)
        at org.apache.knox.gateway.topology.discovery.cm.monitor.ClouderaManagerClusterConfigurationMonitor.<init>(ClouderaManagerClusterConfigurationMonitor.java:100)
        at org.apache.knox.gateway.topology.discovery.cm.monitor.ClouderaManagerClusterConfigurationMonitorProvider.newInstance(ClouderaManagerClusterConfigurationMonitorProvider.java:35)
        at org.apache.knox.gateway.services.topology.impl.DefaultClusterConfigurationMonitorService.init(DefaultClusterConfigurationMonitorService.java:44)
        at org.apache.knox.gateway.services.DefaultGatewayServices.init(DefaultGatewayServices.java:137)
        at org.apache.knox.gateway.GatewayServer.main(GatewayServer.java:184)
        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
        at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.base/java.lang.reflect.Method.invoke(Method.java:566)
        at org.apache.knox.gateway.launcher.Invoker.invokeMainMethod(Invoker.java:68)
        at org.apache.knox.gateway.launcher.Invoker.invoke(Invoker.java:39)
        at org.apache.knox.gateway.launcher.Command.run(Command.java:99)
        at org.apache.knox.gateway.launcher.Launcher.run(Launcher.java:75)
        at org.apache.knox.gateway.launcher.Launcher.main(Launcher.java:52){noformat}
In this particular case, the {{$KNOX_DATA_DIR/cm-clusters/hCM_HOST_7183-Cluster_1.conf}} file was empty so when Knox wanted to load any previously-persisted discovery configuration data into its own in-memory cache, an NPE was thrown.
Although the chance for this to happen is quite low, Knox should handle this situation better.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)