We recently encounter the following problem:
When starting a debug session (e.g. within a test window) we get an ORA-03113 end-of-file on communication channel. Simultanously a trace file is written by the server containing messages like
*** 2003-07-16 15:39:38.842
*** SESSION ID
41.2108) 2003-07-16 15:39:38.832
Probe:read_pipe: receive failed, status 1
Probe:S:debug_loop: timeout. Action 1
*** 2003-07-16 16:03:00.702
Probe:read_pipe: receive failed, status 3
Probe:S:debug_loop: timeout. Action 1
The bad thing on this is that whenever this problem occurs noone will be able to start a debugging session on the database until we shutdown the database (which only works with shutdown abort!) and restart it. We would urgently need a solution to recover from that state without having to restart the DB.
Closing all pipes either by flushing the shared pool or using the appropriate dbms_pipe function doesn't help.
We are using PL/SQL Developer V5.1.0.662 on WIN-NT clients and Oracle Database Server V9.2.0.1.0 under Red Hat Linux
When starting a debug session (e.g. within a test window) we get an ORA-03113 end-of-file on communication channel. Simultanously a trace file is written by the server containing messages like
*** 2003-07-16 15:39:38.842
*** SESSION ID

Probe:read_pipe: receive failed, status 1
Probe:S:debug_loop: timeout. Action 1
*** 2003-07-16 16:03:00.702
Probe:read_pipe: receive failed, status 3
Probe:S:debug_loop: timeout. Action 1
The bad thing on this is that whenever this problem occurs noone will be able to start a debugging session on the database until we shutdown the database (which only works with shutdown abort!) and restart it. We would urgently need a solution to recover from that state without having to restart the DB.
Closing all pipes either by flushing the shared pool or using the appropriate dbms_pipe function doesn't help.
We are using PL/SQL Developer V5.1.0.662 on WIN-NT clients and Oracle Database Server V9.2.0.1.0 under Red Hat Linux