Updated for PAW 2.0.76. Planning Analytics Workspace is IBM’s move of TM1 to have a highly graphical, high-performance visualisation layer on top of TM1. It uses a containerised server (using Docker) and can be installed on either Windows or Linux servers. It does not have to be on the same server as TM1.
We recommend against installing PAW on the TM1 server as you waste PVU licence count on the resources required to run PAW if it is on the TM1 server itself
Windows Server 2016 Only
These instructions are for Windows Server 2016 only. For Windows Server 2019 installation of PAW, please see this post.
Background
The Planning Analytics Workspace server will talk to the TM1 server and directly use and write data to TM1. All cubes, views, dimensions, processes, control objects and applications are available from the TM1 server inside PAW. PAW assets can be migrated from one PAW server to another as required.
There are a number of components to installing and configuring Planning Analytics Workspace on Windows Server 2016. These include:
- Docker
- Docker Compose
- Planning Analytics Workspace
These instructions are for the installation of Planning Analytics Workspace on Windows Server 2016 only. Windows Server 2012 R2 and Linux have different installation procedures.
Docker
Docker Now Installed with PAW
Docker now is installed with PAW. There is no need to do a separate installation of Docker or Docker Compose.
Change Docker Location
However, you probably do not want to have the Docker images cluttering up your OS disk. Therefore we very strongly recommend that you set Docker to use a data disk.
To do this, please do the following:
- Create a folder on a data disk (E, F etc) at the root called “Docker”. Note the casing of the letters as these are important. Our example above has a capital D and the rest lower case.
- Go to C:ProgramData
- Create a folder called “docker”. Lower case “d”.
- Open this new C:ProgramDatadocker
- Create a new folder called “config”.
- In config, create a new text file and paste the following into it:
{
"shutdown-timeout": 180,
"data-root": "f:Docker"
}
- Modify the path to the Docker folder you created above. Mine is on F drive, so my path is f:\Docker. Note the double backslash.
- Save this file as “daemon.json”.
- Now, when you go to install PAW and it installs Docker, the installation will go to ProgramData/docker and read the daemon.json file that redirects the installation to the location you want.
Planning Analytics Workspace
The next component after Docker that needs to be installed is PAW itself. Before we install it though, we need to understand that IBM plans to release a new version of PAW every month or so. Therefore we suggest that you copy the entire unzipped PAW installation folder into a folder that is called something like “PAW” at the root of your data drive. Then it will be easy to access from Windows PowerShell.
Then inside the PAW folder, copy the downloaded installation files from IBM.
Once it’s there, unzip the download and rename the unzipped folder it as “PAW_nnnn” , where the “nnnn” part refers to the version of PAW. So for PAW version 2.0.76 it would be PAW_2076, with the dots removed for ease of use
Then open Windows PowerShell in Administration mode and navigate to the location where you copied the installation files, so something like “F:PAWPAW_2076”.
Then, in PowerShell, run the command:
./Start.ps1
This will then start the process of installing Docker, Docker Compose and then creating the PAW images inside Docker. This will take quite a while to complete (over an hour).
Steps to Install Planning Analytics Workspace
Time needed: 1 hour and 30 minutes.
Here are the detailed steps on how to Install Planning Analytics Workspace as of version 2.0.63. This includes Docker.
- Create a new Windows Server
Create a new Windows server and install all the standard stuff.
- Configure where you want Docker to install its data files
Create a daemon.json file at c:ProgramDatadockerconfig. Please see the steps above for detailed instructions.
- Create a location for Planning Analytics Workspace installation files.
Create a folder at the root of your data drive. Something like FPAWPAW_2076.
Then each time you update, you will add a new folder and migrate from the previous. Please see the detailed steps above for that as well. - Copy the PAW installation files
Copy the downloaded PAW installation files into the new folder, in our example FPAWPAW_2076.
- Run the PAW installer.
1. Open a PowerShell window as an Administrator.
2. Navigate to your PAW folder (like F:PAWPAW_2076).
3. run ./start.ps1
This will initiate installation of Docker, Docker Compose and Planning Analytics Workspace. - Warning for Container Installation
Docker needs Containers to be installed on the Windows server. If it is not installed, the PAW installation process will install it and prompt you to restart the server.
- After Restart, Re-run PAW Installer
1. Open a PowerShell window as an Administrator.
2. Navigate to your PAW folder (like F:PAWPAW_2076).
3. Run ./start.ps1 - Start Planning Analytics Administration tool
Eventually (over an hour later on our server), the image load finished and we were prompted to Start the Administration Tool. Enter Y and hit enter to continue.
- Planning Analytics Workspace Administration Configuration
Hitting Y above will then open Planning Analytics Workspace configuration in your favourite browser.
- Accept Licenses
Accept the IBM and Non IBM licenses.
- Enter Configuration Details
In the Planning Analytics Workspace Administration Tool, copy and paste the various settings from the tm1s.cfg file for your TM1 model.
The values that go in these boxes are as follows:
TM1 Admin Server URI: the URI for your TM1 server, with either HTTPS and port 5898 or HTTP and port 5895
TM1 Application Server URI: Your TM1 Web address WITHOUT the trailing “/tm1web/”
TM1 Login Server URI: The URI from the Admin server above, excluding the port and suffixed with the HTTPPortNumber from tm1s.cfg on your TM1 server.
So if our TM1 server name was “TM1XPA” and our fully qualified address was “tm1xpa.tm1explorers.com”, the HTTPPortNumber from tm1s.cfg was “12345” and we were using HTTPS, then the settings would be likely to be:
https://tm1xpa.tm1explorers.com:5898
http://tm1xpa.tm1explorers.com:9510
https://tm1xpa.tm1explorers.com:12345 - Validate
Now hit Validate at the bottom of the screen. If you have Planning Analytics Spreadsheet Services (PASS) installed, you will no longer be using the application Server Gateway and will get an error like I have below.
Otherwise all checks should pass. - Update
Now hit the Update button and all the PAW services will be updated with the settings you have entered above.
- Wait!
Now is the next time to wait. Flip over to the Status tab and it will probably be blank. Hit the Refresh (not Restart) button and wait.
You can also open Task Manager to see progress. For a new installation of PAW on a fresh Windows Server 2016 box, it will get up to about 14Gb of memory consumed on the Performance tab and CPU will sit at 100% until it is finished. - Launch Planning Analytics Workspace
Open a fresh PowerShell window and navigate to your PAW installation (F:/PAW/PAW_2063).
When there, enter ./scripts/launch.ps1. This will then launch PAW into your browser.
Note that each time PAW restarts, it assigned a new IP address inside the Docker container. Using launch.ps1 will go get the IP address and then open that in your browser. You can’t bookmark that though cos it will be different next time. - Open Planning Analytics Workspace from another PC
Once you know that it is working on the server, go to another PC and enter http://PAWServerName. Because it uses port 80, there is no need to add the port suffix. That should open PAW directly in the browser.
From there you can distribute the address or add it as a connection inside Planning Analytics for Excel for use there as well.
Version Management
Remember that the versions of Planning Analytics Workspace, Planning Analytics for Excel and Planning Analytics Spreadsheet Services all must be kept pretty well in sync. Further, you TM1 (Planning Analytics) server must be on a version that supports these other tools. For more information on them, please see this post.
Need Help?
So that is about it. If you need any help or have any questions about how to install Planning Analytics Workspace or keeping things in sync, please let us know.