Description of problem:
tangd exit code for random get request causes tand.socket@.service to be left in failed state.  So when system is being scanned for vulnerabilities the system can end up having several thousands of these failed services.  These events at a minimum are noise and annoying to administrators.  But if these grow enough, sure can cause delays in systemd.

Version-Release number of selected component (if applicable):
  tang-7-5.el8 
  (also issue on RHEL 7)

How reproducible:
  VERY

Steps to Reproduce:
1. systemctl start tangd.socket
2. echo get / | nc localhost 80
3. systemctl --failed



Actual results:
===============
[root@rhel8-2 ~]# systemctl --failed
  UNIT                                         LOAD   ACTIVE SUB    DESCRIPTION                  
● tangd.0.1:80-127.0.0.1:57160.service loaded failed failed Tang Server (127.0.0.1:57160)

[root@rhel8-2 ~]# systemctl status tangd.0.1:80-127.0.0.1:57160.service 
● tangd.0.1:80-127.0.0.1:57160.service - Tang Server (127.0.0.1:57160)
   Loaded: loaded (/usr/lib/systemd/system/tangd@.service; static; vendor preset: disabled)
   Active: failed (Result: exit-code) since Mon 2020-04-27 15:06:01 CDT; 4min 34s ago
  Process: 48961 ExecStart=/usr/libexec/tangd /var/db/tang (code=exited, status=1/FAILURE)
 Main PID: 48961 (code=exited, status=1/FAILURE)

Apr 27 15:06:01 rhel8-2.localdomain systemd[1]: Started Tang Server (127.0.0.1:57160).
Apr 27 15:06:01 rhel8-2.localdomain systemd[1]: tangd.0.1:80-127.0.0.1:57160.service: Main process exited, code=exited, status=1/FAILURE
Apr 27 15:06:01 rhel8-2.localdomain systemd[1]: tangd.0.1:80-127.0.0.1:57160.service: Failed with result 'exit-code'.



Expected results:
=================
[root@rhel8-2 ~]# systemctl --failed
0 loaded units listed. Pass --all to see loaded but inactive units, too.



Additional info:
================
While this single occurrence is just annoying, on an enterprise subnet that is regularly scanned for security vulnerabilities a system can easily have thousands of these failed entries.

Can of course replicate that easily with:
[root@rhel8-2 ~]# while true ; do echo get / | nc localhost 80 ; done
^C
[root@rhel8-2 ~]# systemctl --failed | grep -c tangd
1714