Student and Proctor Images

TDS Environment Setup/Deployment

DB setup

The assumption is that one has already followed the necessary steps to stand up a DB server covered in previous deployment checklist items. However, there are some new schemas and tables that are required for the deployment. Within the tds_deployment_support_files.zip file there is a db directory with two SQL scripts that can be used to create application tables.

Log into the TDS database
Run the exam_create_schema.sql
- This will create the exam database with the appropriate tables
- The user that you want the application to use should have read and write abilities on these tables.
Run the exam_audit_create_schema.sql
- This will create the exam_audit database with the appropriate tables
- The user that you want the application to use should have read and write abilities on these tables.
The user that you want the application should have the following DB access
- Write access: exam, exam_audit, and session database
- Read access: itembank, configs database

EC2 Container Service

Create EC2 Container Repositories to host Student and Proctor Docker images (separate one for each) with embedded items and stimuli.

S3 Configuration for Items

The item scoring functionality requires items and stimuli to be stored on S3 so that regardless of the number of exam microservices they all share the same item and stimuli. The following section lists the steps to create and set up an S3 bucket. This section does assume that you have an AWS account with the ability to create buckets on s3.

Uploading Items and Stimuli to S3

Create a bucket that will house the items. We also suggest creating a folder within that bucket to isolate the items and stimuli for that deployment.

For example, the bucket can be tds-resources with a folder uat

tds-resources
  uat
    items
      item-187-1234
 item-187-1234.xml
 metadata.xml
    stimuli
       stim-187-1234
  stim-187-1234.xml
  metadata.xml

Student

Clone (git) Student master branch: git clone https://github.com/SmarterApp/TDS_Student.git
Ensure that you are using the proper tag.
Place the items and stimuli that you want to use in this deployment in on the machine you’re using to build the project: /usr/local/tomcat/resources/tds/bank
- Practice and training tests can be downloaded from FTP site: ftp://ftps.smarterbalanced.org/~sbacpublic/Public
Run mvn clean install -DskipTests in the cloned project’s base directory.
Change directory to the student directory cd student
Run mvn docker:build
Tag the docker image that was just built with the Student EC2 Repository you created earlier.
- Run docker images to find the IMAGE ID for what was just built.
- Run docker tag <IMAGE ID> <STUDENT EC2 Repository>:latest
- Example: docker tag 7b6edb762795 269146879732.dkr.ecr.us-west-2.amazonaws.com/student:latest
Run eval $(aws ecr get-login)
Run docker push <STUDENT EC2 Repository>:latest

Proctor

Clone (git) Proctor master branch : git clone https://github.com/SmarterApp/TDS_Proctor.git
Ensure that you are using the proper tag.
This should be done if you already built student. Place the items and stimuli that you want to use in this deployment on the machine you’re using to build this project in /usr/local/tomcat/resources/tds/bank
Change directory to the proctor directory (cd proctor)
Run mvn clean install -DskipTests in the cloned project’s proctor directory.
Get yourself a Proctor SAML xml metadata file and associated keystore *.jks file.
- Notes for this can be found http://www.smarterapp.org/deployment/checklist/art.html
Put the metadata file and keystore in /opt/sbtds/deploy/resources/security
Run mvn docker:build
Tag the docker image that was just built with the Proctor EC2 Repository you created earlier.
- Run docker images to find the IMAGE ID for what was just built.
- Run docker tag <IMAGE ID> <PROCTOR EC2 Repository>:latest
- Example: docker tag 7b6edb762795 269146879732.dkr.ecr.us-west-2.amazonaws.com/proctor:latest
Run eval $(aws ecr get-login)
Run docker push <PROCTOR EC2 Repository>:latest

back to Deployment Checklists