systemd failure during startup Application Server
Article ID: 206814
Updated On:
Products
Issue/Introduction
The application server and scheduler fails on start when issue from /etc/init.d or from unisrvcntr. The problem started after upgrading or installing WAAE 11.3.6 to Linux Red Hat 7.8.
unifstat
CA Services Status Report
Component Name Pid Status
------------------------------------ ------- --------------
WAAE Application Server (DEV) - not active
WAAE Scheduler (DEV) - not active
WAAE Agent (WA_AGENT) 1888 running
CA-diadna Server 2232 running
CA-CCI Server 2502 running
CA-CCI Remote Server 2524 running
CA-CCI Clean Up 2519 running
CA-CCI Legacy Proxy 2522 running
CA-WV Status Server 3276 running
Unicntrl and init.d scripts, thinks it is not running and systemd reports this
[root@test-server bin]# systemctl status waae_server.DEV -l
● waae_server.DEV.service - LSB: CA Workload Automation AE Application Server
Loaded: loaded (/etc/rc.d/init.d/waae_server.DEV; bad; vendor preset: disabled)
Active: failed (Result: exit-code) since Fri 2020-12-18 16:12:22 CST; 30min ago
Docs: man:systemd-sysv-generator(8)
Process: 14005 ExecStart=/etc/rc.d/init.d/waae_server.DEV start (code=exited, status=1/FAILURE)
Dec 18 16:12:17 test-server systemd[1]: Starting LSB: CA Workload Automation AE Application Server...
Dec 18 16:12:22 test-server waae_server.DEV[14005]: WAAE Application Server (DEV) [FAILED]
Dec 18 16:12:22 test-server systemd[1]: waae_server.DEV.service: control process exited, code=exited status=1
Dec 18 16:12:22 test-server systemd[1]: Failed to start LSB: CA Workload Automation AE Application Server.
Dec 18 16:12:22 test-server systemd[1]: Unit waae_server.DEV.service entered failed state.
Dec 18 16:12:22 test-server systemd[1]: waae_server.DEV.service failed.
waae_agent-WA_AGENT.service loaded active running LSB: CA Workload Automation AE Agent
● waae_sched.DEV.service loaded failed failed LSB: CA Workload Automation AE Scheduler
● waae_server.DEV.service loaded failed failed LSB: CA Workload Automation AE Application Server
-.slice loaded active active Root Slice
When attempting to start via systemctl command, this is the message being received.
[root@test-server bin]# systemctl start waae_server.DEV
Job for waae_server.DEV.service failed because the control process exited with error code. See "systemctl status waae_server.DEV.service" and "journalctl -xe" for details.
[root@test-server bin]# systemctl status waae_server.DEV
● waae_server.DEV.service - LSB: CA Workload Automation AE Application Server
Loaded: loaded (/etc/rc.d/init.d/waae_server.DEV; bad; vendor preset: disabled)
Active: failed (Result: exit-code) since Fri 2020-12-18 16:05:44 CST; 29s ago
Docs: man:systemd-sysv-generator(8)
Process: 13498 ExecStart=/etc/rc.d/init.d/waae_server.DEV start (code=exited, status=1/FAIL URE)
Environment
Release : 11.3.6
Component : CA Workload Automation AE (AutoSys)
Cause
The root-cause is the 32-bit AutoSys application is getting crashed in c++ runtime library function named strcasestr due to redhat issue.
Resolution
It is documented in RedHat support at the below URL:
https://access.redhat.com/solutions/3187882
Here is an excerpt from the redhat site:
Issue
A 32 bit binary crashes in a call to strcasestr, but is receiving valid arguments.
Resolution
Update to glibc-2.17-105.el7 or newer.
Create an empty special file to tell glibc to workaround the issue.
Raw
# yum update glibc glibc.i686
# touch /etc/sysconfig/64bit_strstr_via_64bit_strstr_sse2_unaligned
Root Cause
Optimized __strcasestr_sse42 is choosen as strcasestr implementation for 32 bit binaries, but some SSE instructions require 16 byte aligned arguments.
Some old 32 bit binaries, or binaries generated by old environments might not align the stack at 16 byte boundaries, triggering the problem.
Feedback
