Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.

Bug 324934

Summary:	Stale NFS file handle
Product:	Community	Reporter:	Nicolas Bros <nicolas.bros>
Component:	CI-Jenkins	Assignee:	Eclipse Webmaster <webmaster>
Status:	RESOLVED WONTFIX	QA Contact:
Severity:	normal
Priority:	P3
Version:	unspecified
Target Milestone:	---
Hardware:	All
OS:	All
Whiteboard:

Description Nicolas Bros

2010-09-10 05:19:01 EDT

The modisco-integration build on hudson.eclipse.org failed with:

ERROR: IOException encountered while reading "/opt/public/modeling/mdt/modisco/buckminster/install/configuration/org.eclipse.core.runtime/.mainData.24".
java.io.IOException: Stale NFS file handle
	at java.io.RandomAccessFile.readBytes(Native Method)
	at java.io.RandomAccessFile.read(RandomAccessFile.java:315)
	at org.eclipse.core.internal.registry.BufferedRandomInputStream.fillBuffer(BufferedRandomInputStream.java:57)
	at org.eclipse.core.internal.registry.BufferedRandomInputStream.read(BufferedRandomInputStream.java:64)
	at java.io.DataInputStream.readInt(DataInputStream.java:353)
	at org.eclipse.core.internal.registry.TableReader.basicLoadConfigurationElement(TableReader.java:196)
	at org.eclipse.core.internal.registry.TableReader.loadConfigurationElement(TableReader.java:184)
	at org.eclipse.core.internal.registry.RegistryObjectManager.load(RegistryObjectManager.java:411)
	at org.eclipse.core.internal.registry.RegistryObjectManager.basicGetObject(RegistryObjectManager.java:270)
	at org.eclipse.core.internal.registry.RegistryObjectManager.getObject(RegistryObjectManager.java:262)
	at org.eclipse.core.internal.registry.ConfigurationElementHandle.getConfigurationElement(ConfigurationElementHandle.java:26)
	at org.eclipse.core.internal.registry.ConfigurationElementHandle.getAttribute(ConfigurationElementHandle.java:34)
	at org.eclipse.jdt.internal.launching.environments.ExecutionEnvironment.getId(ExecutionEnvironment.java:185)
	at org.eclipse.pde.internal.core.TargetPlatformHelper.getKnownExecutionEnvironments(TargetPlatformHelper.java:304)
	at org.eclipse.pde.internal.core.MinimalState.setExecutionEnvironments(MinimalState.java:254)
	at org.eclipse.pde.internal.core.MinimalState.initializePlatformProperties(MinimalState.java:231)
	at org.eclipse.pde.internal.core.MinimalState.internalResolveState(MinimalState.java:225)
	at org.eclipse.pde.internal.core.MinimalState.resolveState(MinimalState.java:201)
	at org.eclipse.pde.internal.core.PDEState.readTargetState(PDEState.java:126)
	at org.eclipse.pde.internal.core.PDEState.<init>(PDEState.java:100)
	at org.eclipse.pde.internal.core.PDEState.<init>(PDEState.java:92)
	at org.eclipse.pde.internal.core.target.provisional.LoadTargetDefinitionJob.handleReload(LoadTargetDefinitionJob.java:448)
	at org.eclipse.pde.internal.core.target.provisional.LoadTargetDefinitionJob.loadPlugins(LoadTargetDefinitionJob.java:284)
	at org.eclipse.pde.internal.core.target.provisional.LoadTargetDefinitionJob.runInWorkspace(LoadTargetDefinitionJob.java:141)
	at org.eclipse.core.internal.resources.InternalWorkspaceJob.run(InternalWorkspaceJob.java:38)
	at org.eclipse.buckminster.pde.internal.PDETargetPlatform.setTargetActive(PDETargetPlatform.java:143)
	at org.eclipse.buckminster.pde.internal.PDETargetPlatform.refresh(PDETargetPlatform.java:384)
	at org.eclipse.buckminster.pde.internal.PDETargetPlatform.locationsChanged(PDETargetPlatform.java:361)
	at org.eclipse.buckminster.core.materializer.P2Materializer.materialize(P2Materializer.java:326)
	at org.eclipse.buckminster.core.materializer.MaterializerJob.run(MaterializerJob.java:64)
	at org.eclipse.core.internal.jobs.Worker.run(Worker.java:54)

Comment 1 Denis Roy

2010-09-10 10:48:52 EDT

If I am not mistaken, you are running build jobs on build.eclipse.org _and_ hudson.eclipse.org.  If that is correct, since /opt/public is a shared resource between all the Hudson servers, is it possible that one build may have altered/erased a file that the other build was expecting to see?

If I look right now, the file you have mentioned does not exist.

What does that file do?

Comment 2 Nicolas Bros

2010-09-10 11:04:13 EDT

Makes sense! I hadn't thought of that.

> What does that file do?
I don't know exactly, but it is in the Buckminster install configuration, so it is very possible that two instances of Buckminster running concurrently are trying to access this file at the same time, causing conflicts.

Comment 3 Denis Roy

2010-09-10 11:23:38 EDT

> two instances of Buckminster running concurrently are
> trying to access this file at the same time, causing conflicts.

FWIW, access is not the problem.  A Stale NFS file handle would occur under this scenario:

- server A opens file, acquires file handle, but does not request an exclusive lock
- server B deletes file since it's not locked
- server A attempts any file operation on the file, even closing it.  File handle is stale.