You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@knox.apache.org by "Sandor Molnar (Jira)" <ji...@apache.org> on 2023/01/31 12:42:00 UTC
[jira] [Work started] (KNOX-2869) Possible NPE at CM cluster configuration monitor startup
[ https://issues.apache.org/jira/browse/KNOX-2869?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Work on KNOX-2869 started by Sandor Molnar.
-------------------------------------------
> 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
> Priority: Major
> 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)