hansva opened a new issue, #8455: URL: https://github.com/apache/hop/issues/8455
### Apache Hop version? 2.20.0-SNAPSHOT ### Java version? 21 ### Operating system Other ### What happened? Since the per-execution VFS namespaces (#8106, #8231), a Hop Server can deadlock when a second exported workflow/pipeline arrives while another one is loading its files. The stuck execution never finishes, the client's RemoteWorkflowEngine polls it forever, and the server stays otherwise alive. The load-balance integration test (0002-drive-100-workflows) hits this in every nightly since #2311; the last server log line is Starting action [0001-child-delay.hpl] and hop-run is killed by the watchdog after 3600 s. Reproduced locally with run-tests-docker.sh PROJECT_NAME=load-balance; a kill -3 on the server JVM reports a Java-level deadlock: Thread A (running workflow, loading its child pipeline): HopVfs.getFileObject → synchronized HopVfs.getFileSystemManager(variables) (holds HopVfs.class) → HopVfsNamespaces.resolve → synchronized HopVfsNamespaces.existing (waits for HopVfsNamespaces.class) Thread B (second workflow starting, Workflow.startExecution): synchronized HopVfsNamespaces.acquire (holds HopVfsNamespaces.class) → HopVfsNamespace.registerNamedProviders → AzureVfsPlugin.getProviders → JsonMetadataSerializer.validateBaseFolder → synchronized HopVfs.getFileObject (waits for HopVfs.class) ### Issue Priority Priority: 2 ### Issue Component Component: Hop Server -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
