-
Bug
-
Resolution: Unresolved
-
Critical
-
None
-
rhel-9.7
-
None
-
No
-
Important
-
rhel-sst-cs-stacks
-
ssg_core_services
-
None
-
False
-
-
None
-
None
-
None
-
None
-
-
All
-
None
What were you trying to do that didn't work?
An IDM server is becoming unresponsive after some time and begins logging:
[Wed Oct 30 17:14:37.500311 2024] [mpm_event:error] [pid 655880:tid 655880] AH03490: scoreboard is full, not at MaxRequestWorkers.Increase ServerLimit.
All child processes are created and pstacks show they all are stuck in an attempted exit like below:
#0 0x000003ffa649ab26 in __futex_abstimed_wait_cancelable64 () from target:/lib64/libc.so.6 #1 0x000003ffa64a63d8 in do_futex_wait.constprop () from target:/lib64/libc.so.6 #2 0x000003ffa64a6478 in __new_sem_wait_slow64.constprop.0 () from target:/lib64/libc.so.6 #3 0x000003ffa1baafba in PyThread_acquire_lock_timed () from target:/lib64/libpython3.9.so.1.0 #4 0x000003ffa1c65426 in acquire_timed () from target:/lib64/libpython3.9.so.1.0 #5 0x000003ffa1c651be in lock_PyThread_acquire_lock () from target:/lib64/libpython3.9.so.1.0 #6 0x000003ffa1bd3132 in method_vectorcall_VARARGS_KEYWORDS () from target:/lib64/libpython3.9.so.1.0 #7 0x000003ffa1bc0e8c in _PyEval_EvalFrameDefault () from target:/lib64/libpython3.9.so.1.0 #8 0x000003ffa1bd1e24 in function_code_fastcall () from target:/lib64/libpython3.9.so.1.0 #9 0x000003ffa1bd19e8 in _PyFunction_Vectorcall () from target:/lib64/libpython3.9.so.1.0 #10 0x000003ffa2011f48 in ShutdownInterpreter_call () from target:/etc/httpd/modules/mod_wsgi_python3.so #11 0x000003ffa1bcb832 in _PyObject_MakeTpCall () from target:/lib64/libpython3.9.so.1.0 #12 0x000003ffa1be8604 in PyObject_VectorcallMethod () from target:/lib64/libpython3.9.so.1.0 #13 0x000003ffa1ca13a4 in wait_for_thread_shutdown () from target:/lib64/libpython3.9.so.1.0 #14 0x000003ffa1b9e072 in Py_EndInterpreter () from target:/lib64/libpython3.9.so.1.0 #15 0x000003ffa2010a7e in Interpreter_dealloc () from target:/etc/httpd/modules/mod_wsgi_python3.so #16 0x000003ffa1bbc1d8 in free_keys_object () from target:/lib64/libpython3.9.so.1.0 #17 0x000003ffa2019e4c in wsgi_python_child_cleanup () from target:/etc/httpd/modules/mod_wsgi_python3.so #18 0x000003ffa66ad052 in apr_pool_destroy () from target:/lib64/libapr-1.so.0 #19 0x000003ffa3e863f8 in clean_child_exit () from target:/etc/httpd/modules/mod_mpm_event.so #20 0x000003ffa3e89180 in child_main () from target:/etc/httpd/modules/mod_mpm_event.so #21 0x000003ffa3e89744 in make_child () from target:/etc/httpd/modules/mod_mpm_event.so #22 0x000003ffa3e89806 in startup_children () from target:/etc/httpd/modules/mod_mpm_event.so #23 0x000003ffa3e8a8f4 in event_run () from target:/etc/httpd/modules/mod_mpm_event.so #24 0x000002aa39cae3ca in ap_run_mpm () #25 0x000002aa39c996d8 in main ()
So this seems to resemble https://github.com/GrahamDumpleton/mod_wsgi/issues/730.
What is the impact of this issue to you?
httpd eventually becomes unresponsive till restarted
Please provide the package NVR for which the bug is seen:
httpd-2.4.57-5.el9.s390x
python3-mod_wsgi-4.7.1-11.el9.s390x
Expected results
httpd becomes unresponsive across any reloads or any other source of process destruction attempts (like mpm pool adjustments).
Actual results
httpd does not become unesponsive.
