|Pivotal GemFire||7.x - 126.96.36.199|
This article discusses how to resolve a gfsh hang issue that may occur when running [gfsh start]/[gfsh status].
When using the [gfsh start] command to start a locator/cacheserver or when using the [gfsh status] command to confirm GemFire member process status, the gfsh shell may not return a result, or may simply continue showing [.......] in the console, while the GemFire member process does start correctly or is in a healthy status.
There is an underlying bug (GEM-809/GEODE-1387) involving a race condition in the file-based support for the [gfsh start] and [gfsh status] commands.
With these two commands, if the JDK's tools.jar is not in the CLASSPATH, gfsh will use a file-based implementation for [gfsh start] and [gfsh status] and may encounter a hang issue during [gfsh start] or [gfsh status] execution. (The [gfsh start] command uses the [gfsh status] routine to determine if startup is complete.)
When the tools.jar is in the CLASSPATH, the [gfsh start]/[gfsh status] commands will use the tools.jar's API which will not hit this hang issue.
This issue is fixed with GemFire 188.8.131.52+/Gemfire9.x. and may be avoided/worked around in earlier versions by adding the tools.jar to the CLASSPATH of gfsh and any locator/server.
For further information, please refer to the following resources: