How to remedy Polyspace "Error: No QueueManager server found."

I have recently moved our R2013a license to a new Linux/x86_64 machine. License installation went OK (after a few hickups with CentOS 7 calling eth0 now enp8s0, which I undid).

I can submit Polyspace jobs, they get processed and the results are saved in the directory specified as datadir in /etc/Polyspace/polyspace.conf, i.e.

# ls -l /opt/PolyspaceResults
-rwxr-x--- 1 daemon daemon    7463 Apr 25 13:22 Custom-SQO-Definitions.xml
-rwxr-x--- 1 daemon daemon 1224515 Apr 25 13:26 analysis1-downloaded.zip
-rwxr-x--- 1 daemon daemon  167756 Apr 25 13:26 analysis1-log.txt
-rwxr-x--- 1 daemon daemon     219 Apr 25 13:26 analysis1-progress.txt
-rwxr-x--- 1 daemon daemon 1335279 Apr 25 13:24 analysis1.zip
drwxrwxrwx 2 daemon daemon       3 Apr 25 13:04 locks
-rwxr-x--- 1 daemon daemon     762 Apr 25 13:22 polyspaced-log-2019-04-25.2.zip
-rwxr-x--- 1 daemon daemon    1257 Apr 25 14:24 polyspaced-log-2019-04-25.3.zip
-rwxr-x--- 1 daemon daemon    1014 Apr 25 14:58 polyspaced-log-2019-04-25.4.zip
-rwxr-x--- 1 daemon daemon     128 Apr 25 13:04 polyspaced-log-2019-04-25.zip
-rw-r--r-- 1 daemon daemon    1124 Apr 25 14:58 polyspaced.log
-rwxr-x--- 1 daemon daemon     334 Apr 25 13:26 queue_infos.txt
drwxr-x--- 2 daemon daemon       5 Apr 25 13:22 results-repository
drwxr-x--- 6 daemon daemon       6 Apr 25 14:58 tomcat

However, when I want to start the polyspace-spooler (on the server machine) or other commands to look at the results, I get after a 10 second timeout:

# /opt/PolyspaceForCandCPP_R2013a/polyspace/bin/psqueue-dump -f
Error: No QueueManager server found.

The polyspaced is running, though (and it does process my verification jobs).

What do I have to do so contacting the QueueManager works? Did I botch my installation?

The polyspaced.log has this to say (looks good):

[...]
User: daemon
Checking files locations...
- data folder (/opt/PolyspaceResults)
- configuration file (/etc/Polyspace/polyspace.conf)
- locks folder (/opt/PolyspaceResults/locks)
- Results repository directory (/opt/PolyspaceResults/results-repository)
- log file (/opt/PolyspaceResults/polyspaced.log)
***********************************
* Checking CPUs configuration... *
***********************************
Checking configuration for CPU marvin...
Processor : AMD Ryzen Threadripper 2950X 16-Core Processor
R.A.M. : 62.70 GB
O.S. : CentOS Linux release 7.6.1810 (Core)
User: daemon
- product 1 :
  - installation folder : /opt/PolyspaceForCandCPP_R2013a/polyspace
  - version : 8.5 (R2013a)
- supported languages : CPP C
The CPU is free (1/1 licenses available).
=====================================
=> The configuration is now correct.
=====================================
26-Apr-2019, 10:36:17: Initializing connections...
26-Apr-2019, 10:36:17: Listening to port 12427...

Answer 1

[Jens Schweikhardt]

The missing piece was the firewall configuration. The new machine had the firewalld running with only a few ports open.

To configure the firewalld of CentOS 7, punching a hole for UDP port 12427, I ran these commands as root:

# firewall-cmd --permanent --add-port=12427/udp
success
# firewall-cmd --reload
success
# /opt/PolyspaceForCandCPP_R2013a/polyspace/bin/psqueue-dump -f
ID      GROUP_ID        AUTHOR  APPLICATION     LOCAL_RESULTS_DIR       CPU     STATUS  DATE    LANG
1       0       [...]

Yay!