Process Scheduler Auto Update

I have been setting up process monitor auto update, and have managed to get it to work - I can see the process status updates in the process monitor screen. Here is what I did.

Oracle support document id 2772617.1 explains how to set this up manually. I wanted this to work as part of the automated build, which means supplying the parameters as part of psft_customisations.yml

A Gotcha!

We have multiple application servers running on the same port (but on different VMs). This means we need different domain IDs for each process scheduler because domain ID and port are used as a key. I would have thought that the hostname should be included to make this a unique identifier, but Oracle have chosen not to do that. Note that domain ID is different from domain name. Oracle documentation suggests using the database name in lower case. The DPK default is APPDOM (which is also the default domain name). If either of these is used, when you set up the inter domain event credentials on the process scheduler and configure the domain (for example by running):

1

psadmin -p configure -d PRCSDOM

the following error messages will appear (wrapped for readability):

1
2
3
4
5

CMDGW_CAT:1521: ERROR: Invalid parameter:
    Entry in DM_REMOTE (AS_APPDOM_9000) appears more than once
CMDGW_CAT:1522: ERROR: Invalid parameter:
    Entry in DOMAIN_ID (APPDOM_9000) appears more than once
CMDGW_CAT:1546: ERROR: dmloadcf: Above errors found during syntax checking

Setting Domain IDs

Application Server

So I have to think of a way to make the domain id unique. We name our VMs with a number on the end, so I use the last character of the hostname together with the first seven characters of the environment name. In our Jinja template this ends up as:

1
2

      Domain Settings/Domain ID:    "{{ environment_name[:7] | lower }}
        {{- inventory_hostname_short[-1] }}"

We set environment_name in our Ansible inventory, and inventory_hostname_short is set by ansible. Of course I need to remember that this won’t work if we have more than 10 application server VMs.

As noted above, the domain name can be different to the domain ID, so we keep a consistent name to ensure that the paths are the same on each VM.

Process Scheduler

We only have two process scheduler VMs, one on Unix and one on Windows, so these have hardcoded domain IDs in the psft_customizations.yml

Setting the Inter-Domain Event Credentials

Credential String Format

It is these domain IDs which need to go into the credential string. It appears to be the following format for the process schedulers:

1
2

PRCS_{{ domain_id }}_{{ prcs_notification_port }}|
    {{- hostname }}:{{ prcs_notification_port }}

e.g.

PRCS_PSUNX_8988|testpsunx.example.com:8988

And the following for the application serves:

1

{{ domain_id }}_{{ apps_listen_port }}|{{ hostname }}:{{ apps_notification_port }}

e.g.

testenv1_9000|testapps1.example.com:7988

Remember this credential string is what we are talking to, so we need to specify the application server hosts in the process scheduler configuration, and vice versa.

In the Process Scheduler Customisations YAML file

In the psft_customizations.yaml jinja template for ansible, we need to write something like this:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16

      Inter-Domain Events/Application Server Credentials: "
      {#- Whitespace control means this should all end up on one line -#}
      {%- for server in groups['app_servers'] -%}
         {{- environment_name[:7] }}
         {{- server.split('.')[0][-1] -}}
         _
         {{- apps_listen_port -}}
         |
         {{- server -}}
         :
         {{- apps_notification_port }}
         {%- if not loop.last -%}
           ,
         {%- endif -%}
      {%- endfor -%}
      "

So we are looping through the application servers, and filling in the environment name, the ports and the server names. Each credential is separated by a comma, but we don’t want a comma at the end of the line.

it generates the string (wrapped for readability).

1
2

      Inter-Domain Events/Application Server Credentials:
        "testenv1_9000|testapps1:7988,testenv2_9000|testapps2:7988"

In the Application Server Customisations YAML file

Since in our setup we always have a Unix process scheduler, and normally also a Windows one, I hardcode that information:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

      Inter-Domain Events/Process Scheduler Credentials: PRCS_PSUNX_8988|
        {{- groups['unix_process_servers'][0] -}}
        :
        {{- prcs_notification_port }}
        {%- if 'windows_process_servers' in groups -%}
          ,PRCS_PSNT_
          {{- prcs_notification_port -}}
          |
          {{- groups['windows_process_servers'][0] -}}
          :
          {{- prcs_notification_port -}}
        {% endif %}

This renders the line like this (Wrapped for readability):

1
2

      Inter-Domain Events/Process Scheduler Credentials:
        PRCS_PSUNX_8988|:8988,PRCS_PSNT_8988|:8988

Checking it Out

I have saved the best till last. It is really useful to check snippets of jinja configuration like this. As I write this there is a useful jinja test parser. It’s source is on GitHub. Note that it does appear to post the contents to the server, so don’t put anyting sensitive into it! Paste the following into the Values YAML box (to simulate the ansible inventory):

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16

groups:
  app_servers:
    - testapps1
    - testapps2
  unix_process_servers:
    - testpsunx
  windows_process_servers:
    - testpsnt

inventory_hostname_short: hostname1
hostname: hostname1.example.com
environment_name: testenv
apps_listen_port: 9000
apps_notification_port: 7988
prcs_notification_port: 8988
domain_id: domain

Then any of the snippets above can be pasted into the template box, and the template will be rendered! Try changing the names, the number of application servers etc. to see the effect! How amazing is that!

Troubleshooting

Windows

I had issues with Windows - The application server could not connect to the process scheduler, despite them both being in the same VLAN (i.e., there was no firewall between then to block traffic)

The following error (wrapped for readability) was written to the application server TUXLOG file:

LIBGWT_CAT:1249: WARN: Connection to PRCS_PRCS_PSNT_8988
      address //testpsnt:8988 failed, Network error(115)
LIBGWT_CAT:1304: WARN: No more remote domain address for remote domain
      PRCS_PRCS_PSNT_8988

I remembered that Windows has it’s own firewall (So does Linux in fact) and it turned out that this is what was blocking access. Windows doesn’t have an equivalent to netcal -l, but stack overflow provides a powershell script that can be used to test:

1
2
3
4
5
6
7
8
9

$Listener = [System.Net.Sockets.TcpListener]8988;
$Listener.Start();
while($true) 
{
    $client = $Listener.AcceptTcpClient();
    Write-Host "Connected!";
    $client.Close();
}
$Listener.Stop();

I used Ansible to create a firewall rule to open port 8988:

1
2
3
4
5
6
7
8
9

- name: Open the windows port so that auto update works
  community.windows.win_firewall_rule:
    name: PeopleSoft Process Scheduler Notification Port
    localport: 8988
    action: allow
    direction: in
    protocol: tcp
    state: present
    enabled: true

Annoying errors

I also received some errors in the application server TUXLOG. It turned out that I had configured a process scheduler to talk to the application server, but hadn’t configured the application server to talk to the process scheduler. Thus the application server kept rejecting the connection attempts from the process scheduler.

Again the messages are wrapped for readability:

113631.testapps1!GWTDOMAIN.76250.2848200448.0: LIBGWT_CAT:1509:
    ERROR: Error occurred during security negotiation - closing connection
113730.testapps1!GWTDOMAIN.76250.2848200448.0: LIBGWT_CAT:1073:
    ERROR: Unable to obtain remote domain id (PRCS_PSUNX_8988)
        information from shared memory

Everything is set up OK and it Still is not working!

Try clearing the caches! Including the web server cache.

Conclusion

Despite what it would appear from Oracle’s documentation, it is possible to automate the setup of the process scheduler auto update. This will save us a lot of time in setting up our environments.