-
Bug
-
Resolution: Fixed
-
P3
-
1.1.6
-
b01
-
sparc
-
solaris_2.5
From RMI-USERS list:
We are running into an intermittent problem with an RMI client application.
About 20% of the time, when making remote calls, we get the following
exception.
*** Security Exception: threadgroup ***
java.rmi.RMISecurityException: security.threadgroup
at
java.rmi.RMISecurityManager.checkAccess(RMISecurityManager.java:92)
at java.lang.ThreadGroup.checkAccess(ThreadGroup.java)
at java.lang.Thread.init(Thread.java)
at java.lang.Thread.<init>(Thread.java)
at sun.rmi.transport.RMIThread.newThread(RMIThread.java:45)
at sun.rmi.transport.tcp.TCPChannel.free(TCPChannel.java:222)
at sun.rmi.server.UnicastRef.done(UnicastRef.java:153)
at Server_Stub.getRecord(Server_Stub.java:136)
at Eml.openRecord(Eml.java:369)
at Eml.openOtherRecord(Eml.java:348)
at EmlWidget.actionPerformed(Eml.java:686)
:
:
:
It looks as if the stub class is calling something that leads to an attempt
to create a thread. The RMISecurityManager does not allow downloaded code
to manipulate threads, so this causes an exception to be thrown. So the
question is why is a stub trying to create a thread when this is guaranteed
to throw an exception? Also, why is the problem intermittent?
This is occurring on Solaris 2.6 with JDK 1.1.6. I was unable to duplicate
the problem using JDK 1.1.5.
We are running into an intermittent problem with an RMI client application.
About 20% of the time, when making remote calls, we get the following
exception.
*** Security Exception: threadgroup ***
java.rmi.RMISecurityException: security.threadgroup
at
java.rmi.RMISecurityManager.checkAccess(RMISecurityManager.java:92)
at java.lang.ThreadGroup.checkAccess(ThreadGroup.java)
at java.lang.Thread.init(Thread.java)
at java.lang.Thread.<init>(Thread.java)
at sun.rmi.transport.RMIThread.newThread(RMIThread.java:45)
at sun.rmi.transport.tcp.TCPChannel.free(TCPChannel.java:222)
at sun.rmi.server.UnicastRef.done(UnicastRef.java:153)
at Server_Stub.getRecord(Server_Stub.java:136)
at Eml.openRecord(Eml.java:369)
at Eml.openOtherRecord(Eml.java:348)
at EmlWidget.actionPerformed(Eml.java:686)
:
:
:
It looks as if the stub class is calling something that leads to an attempt
to create a thread. The RMISecurityManager does not allow downloaded code
to manipulate threads, so this causes an exception to be thrown. So the
question is why is a stub trying to create a thread when this is guaranteed
to throw an exception? Also, why is the problem intermittent?
This is occurring on Solaris 2.6 with JDK 1.1.6. I was unable to duplicate
the problem using JDK 1.1.5.