A Prometheus exporter of metrics from Deltek Workbook
Metrics are exported on port 9701 by default. They are available at http://example.com:9701/metrics.
In my experience. Workbook is very slow. Scrapes can exceed 60 seconds. Especially, if you have more than one company.
Install dependencies with pip pip install prometheus_client workbook_api
workbook_employees_profit_percent_bucket{company_id="",le=""} workbook_employees_profit_percent_count{company_id=""} workbook_employees_profit_percent_sum{company_id=""}
workbook_employees_hours_sale_bucket{company_id="",le=""} workbook_employees_hours_sale_count{company_id=""} workbook_employees_hours_sale_sum{company_id=""}
workbook_employees_hours_cost_bucket{company_id="",le=""} workbook_employees_hours_cost_count{company_id=""} workbook_employees_hours_cost_sum{company_id=""} 0.0
workbook_employees_days_employed_bucket{company_id="",le=""} workbook_employees_days_employed_count{company_id=""} workbook_employees_days_employed_sum{company_id=""}
workbook_jobs_days_old_bucket{billable="",company_id="",le=""} workbook_jobs_days_old_count{billable="",company_id=""} workbook_jobs_days_old_sum{billable="",company_id=""}
workbook_billable_jobs{company_id=""}
workbook_jobs_total{company_id=""}
JSON files with dashboards to graph the Prometheus metrics can be found in
the dir grafana.
A dashboard with stats for projects (Jobs in Workbook terminology) is in
the file dashboard_workbook_projects.json. Use the import feature in Grafana
to start using it.
After import in Grafana, change variables company_id and wb_job to match the
labels company_id and job in your Prometheus metrics. When saving the
dashboard after setting the variables, make sure that you enable the
option "save current variables".
A dashboard with financial data is in
the file dashboard_workbook_finace.json. Use the import feature in Grafana
to start using it.
After import in Grafana, change variables company_id and wb_job to match the
labels company_id and job in your Prometheus metrics. When saving the
dashboard after setting the variables, make sure that you enable the
option "save current variables".
The dashboards has the following variables which can be changed by the user
- "Finance accounts": The accounts to show graphs for (Multi choice)
- "Days to use when calculating change"
- "Number of days to base prediction on"
- "Number of days in the future to predict for"
The dashboard includes the following graphs
- Account balance
- Account prediction
- Account balance change
- Debit & Credit
- Debitors & Creditors
- Employment length
- Departments: Sum of hourly costs
- Departments: Average hourly cost
Run the workbook_exporter in a docker container
Run workbook_exporter in a container passing credentials as environment variables:
docker run --detach --name wbe -p 9701:9701 -e "WORKBOOK_URL=https://your.url/api" -e "WORKBOOK_USER=user" -e "WORKBOOK_PASSWORD=secret" tobiasbp/workbook_exporter
Anyone can scrape your data, so you should not allow incomming traffic by exposing the port, unless you are on a trusted network.
Use the docker-compose with the file docker-compose.yml to set up the workbook exporter
behind an nginx reverse proxy with authentication and Let's Encrypt certificates.
You can edit the file, if you do not want to configure the setup using environment
variables, as I have done in the following.
Make sure you have the apache2-utils/httpd-tools, so you can create a password file for nginx to use.
Set environment variables:
- export WORKBOOK_URL="https://me.workbook.dk/api"
- export WORKBOOK_USER="my_user"
- export WORKBOOK_PASSWORD="secret"
- export LETSENCRYPT_MAIL="[email protected]"
- export WORKBOOK_VIRTUAL_HOST="wbe.example.com"
Make sure $WORKBOOK_VIRTUAL_HOST resolves to the IP running your exporter. Otherwise, the certificates can not be issued.
Create a directory to hold the password file for Nginx to use:
mkdir ~/htpasswd
Create password file for user scrape_user. This is the user Prometheus should use when scraping:
htpasswd -c ~/htpasswd/$WORKBOOK_VIRTUAL_HOST scrape_user
You should now have a password file with the name of your virtual host in ~/htpasswd.
Start the containers:
docker-compose up --detach [docker-compose.yml]
Let's make sure everything is doing what we expect:
You should be asked for credentials (Status code 401) when connecting on ports 9701 & 443:
curl https://wbe.everland.dk:9701curl https://wbe.everland.dk
Port 80 should be forwarded (Status code 301):
curl http://wbe.everland.dk
You should get the Workbook metrics when logging in with your credentials:
curl --user scrape_user https://wbe.everland.dk/metrics
Create a new namespace
kubectl create namespace workbook-exporter
Create a secret with the Workbook credentials
kubectl create secret --namespace workbook-exporter generic wbe-secret --from-literal=workbook_url='https://example.workbook.dk/api' --from-literal=workbook_user='my_user' --from-literal=workbook_password='my_password'
Edit wbe-configmap.yml and apply if you want non-default config
kubectl apply -f wbe-configmap.yml
Add workbook-exporter as a service to your cluster
kubectl apply -f wbe-service.yml
Stuff to do. Should this not be in issues in GitHub?
- Status of active jobs in buckets