• Abstract
• Installation
  • Docker
  • Kubernetes
  • Helm
  • From Source
• Postee Configuration File
  • General Settings
  • Routes
    • Route Plugins
  • Templates
  • Outputs
    • ServiceNow
    • Jira
    • Email
    • Slack
    • MsTeams
    • Splunk
    • Generic Webhook
• Configure the Aqua Server with Webhook Integration
• Customizing Templates
• Postee UI
• Misc

Postee is a simple message routing application that receives JSON input messages through a webhook interface, and delivers them based on rules to a set of collaboration systems, including: JIRA, Email, Slack, Microsoft Teams, ServiceNow, Splunk and Generic WebHook.

Primary use of Postee is to act as a notification component for Aqua Security products. It's extremely useful for sending vulnerability scan results or audit alerts from Aqua Platform to collaboration systems.

To run Postee you will first need to configure the Postee Configuration File, which contains all the message routing logic. After the configuration file is ready, you can run the official Postee container image: aquasec/postee:latest, or compile it from source.

There are different options to mount your customize configuration file to Postee - if running as a Docker container, then you simply mount the configuration files as a volume mount. If running as a Kubernetes deployment, you will need to mount it as a ConfigMap. See the below usage examples for how to run Postee on different scenarios.

After Postee will run, it will expose two endpoints, HTTP and HTTPS. You can send your JSON messages to these endpoints, where they will be delivered to their target system based on the defined rules.

To run Postee as a Docker container, you mount the cfg.yaml to '/config/cfg.yaml' path in the Postee container.

docker run -d --name=postee -v /<path to configuration file>/cfg.yaml:/config/cfg.yaml \
    -e POSTEE_CFG=/config/cfg.yaml -e POSTEE_HTTP=0.0.0.0:8084 -e POSTEE_HTTPS=0.0.0.0:8444 \ 
    -p 8084:8084 -p 8444:8444 aquasec/postee:latest

When running Postee on Kubernetes, the configuration file is passed as a ConfigMap that is mounted to the Postee pod.

See Kubernetes instructions to run Postee on Kubernetes using deployment yamls.

When running Postee on Kubernetes, the configuration file is passed as a ConfigMap that is mounted to the Postee pod.

See Helm instructions to run Postee on Kubernetes using Helm chart.

Clone and build the project:

git clone [email protected]:aquasecurity/postee.git
make build

After that, modify the cfg.yaml file and set the 'POSTEE_CFG' environment variable to point to it.

export POSTEE_CFG=<path to cfg.yaml>
./bin/postee

When Postee receives a message it will process it based on routing rules and send it to the appropriate target. How does it know how to do that? Well, this information is defined in Postee's configuration file, cfg.yaml, which contains the following definitions:

1. General settings
2. Routes
3. Templates
4. Outputs

NOTE See examples of common Postee configuration files here: Examples

General settings are specified at the root level of cfg.yaml. They include general configuration that applies to the Postee application.

A route is used to control message flows. Each route includes the input message condition, the template that should be used to format the message, and the output(s) that the message should be delivered to.

The most important part of a route is the input definition using the Rego language to define what are the conditions for an incoming message to be handled by a certain route.

NOTE See the complete Rego Language in OPA-reference

After defining the route's input condition, what is left is to define the template that will be used to format the input message, and the output that formatted message will be sent to.

The below table describes the fields to define a route:

The rego-filters folder contains examples of policy related functions. You can use the examples. To do this, you need to change the input data in the arrays of rego files and fill in the config file. If you want to use an other folder, set the 'REGO_FILTERS_PATH' environment variable to point to it. When using 2 or more files, they will be combined by "OR". To combine policy related functions by "AND", use the Policy-Related-Features.rego file, change the input data, and fill in the required function in allow.

allow{
    PermitImageNames
    PermitMinVulnerability
}

If you are using your own rego files, then the package field should be "postee" and the result should be in the allow function:

package postee

your_function{...} # 0 or more your functions

allow {
    your_function
}

For example, the following input definition will match JSON messages that have 'image.name' field with value that contains the string 'alpine':

input: contains(input.image,"alpine")

Another example using regular expression:

input: regex.match("alp:*", input.image)

You can create more complex input definitions using the Rego language. For example, the following input definition will match JSON messages that have 'image.name' field with value 'alpine' and that their registry is 'Docker Hub' and they have a critical vulnerability.

input: |
  contains(input.image,"alpine")
  contains(input.registry, "Docker Hub")
  input.vulnerability_summary.critical>0

NOTE See more route samples configuration HERE

'Plugins' section contains configuration for useful Postee features.

Templates are used to format input messages before sending them to the output. For example - before sending a message to Microsoft Teams there is a need to format the input JSON into an HTML. This is done using a template.

Each template has a 'name' field, which is used by the route to assign the template to input and output. In addition to name, a template will have one of the 4 below keys:

Outputs are remote services that messages should be sent to. Each output has two mandatory fields, which are 'name' and 'type'.

Depending on the 'type', additional parameters are required.

Follow these steps to set up JIRA integration:

1. Get a new token for user:
   • Login to Jira Cloud. Go to the user profile API tokens (JIRA Cloud users can find it here: https://id.atlassian.com/manage-profile/security/api-tokens). Click on the Create API Token. A new API token for the user is created.
   • Login to Jira Server/Data center Select your profile picture at top right of the screen, then choose Settings > Personal Access Tokens. Select Create token. Give your new token a name. Optionally, for security reasons, you can set your token to automatically expire after a set number of days. Click Create. A new PAT for the user is created.
2. Fill jira output in cfg.yaml:
   • Jira Cloud:
     • User: your email.
     • Password: your API token.
   • Jira Server/Data center:
     • User: your UserName.
     • Password: your Password.
       or
     • Token: your Personal Access Tokens.

For Jira you can also specify custom fields that will be populated with values. Use the unknowns parameter in cfg.yaml for custom fields. Under the unknowns parameter, specify the list of fields names to provide value for. You can add "-numeric-field", "-multiple-value", "multiple-line-text-field", "-date-time-picker" and "-field-url" as suffix to the custom field name, to specify what is the field type.

For example:

unknowns:
     mycustom: "this is a text custom field"
     mycustom-numeric-field: 123
     mycustom-multiple-value: 1,2,3 
     mycustom-multiple-line-text-field: "text \n moretext" 
     mycustom-date-time-picker: 2014-04-11T12:14:26.880+0400
     mycustom-url: https://tour.golang.org/moretypes/7

Getting the Slack webhooks Create a Slack Custom App.

Copy webhook url to the Postee config

Open your Microsoft Teams client. Click on the "..." near the channel you would like to send notifications to.

Choose "Connectors". The connectors window will open. Look for the "Incoming Webhook" connector (it is under the "All" category).

Click "Add" near the Incoming Webhook connector. Click "Add" again. Provide a name and click "Create".

You will be provided with a URL address. Copy this URL and put it in the cfg.yaml.

You will need to care about an HTTP Event Collector in Splunk Enterprise or Splunk Cloud.

This can usually be found in the Splunk console under "Settings -> Data Inputs -> HTTP Event Collector -> Add New".

Once you create an HTTP Event Collector you will receive a token. You should provide this token, together with the Splunk HTTP Collector URL, as part of the cfg.yaml settings.

Postee can be integrated with Aqua Console to deliver vulnerability and audit messages to target systems.

You can configure the Aqua Server to send a Webhook notification whenever a new vulnerability is found. Navigate to the Image Scan Results Webhook page, under the "Settings" menu.

Click "Enable sending image scan results to webhook", and specify the URL of Postee. Now, scan an image and look at the Postee log files - you will see that Postee have received an incoming message once scan is done, and that the message was routed based on the cfg.yaml configuration.

You can also configure the Aqua Server to send a Webhook notification for every audit message. Navigate to the Log Management page, under the "Integrations" menu.

Click on the "Webhook" item, and specify the URL of Postee.

Now every audit event in Aqua will be sent to Postee. You can configure routes and input message conditions in Postee cfg.yaml to forward appropriate messages to target systems.

The URL is in the following formats: HTTPS: https://:8445 or HTTP: https://:8082

To validate that the integration is working, you can scan a new image for security vulnerabilities from the Aqua Server UI (Images > Add Image > Specify Image Name > Add).

When vulnerabilities are found in an image, you will see that a JIRA ticket is created/ Email is received/ Slack message is posted to the channel.

To troubleshoot the integration, you can look at both the Aqua Postee container logs and the Aqua Server logs. Use the "docker logs " command to view these logs.*

Postee loads bundle of templates from rego-templates folder. This folder includes several templates shipped with Postee, which can be used out of the box. You can add additional custom templates by placing Rego file under the 'rego-templates' directory.

To create your own template, you should create a new file under the 'rego-templates' directory, and use the Rego language for the actual template code.

Message payload is referenced as input when template is rendered. The result variable should be used to store the output message, which is the result of the template formatting.

The following variables should be defined in the custom Rego template.

So the simplest example of Rego template would look like:

package example.vuls.html

title:="Vulnerabilities are found"
result:=sprintf("Vulnerabilities are found while scanning of image: <i>%s</i>", [input.image])

Two examples are shipped with the app. One produces output for slack integration and another one builds html output which can be used across several integrations. These example can be used as starting point for message customization

Postee provides a simple Web UI to simplify the configuration management.

See Postee UI for details how to setup the Postee UI.

The Postee container uses BoltDB to store information about previously scanned images. This is used to prevent resending messages that were already sent before. The size of the database can grow over time. Every image that is saved in the database uses 20K of storage.

Postee supports ‘PATH_TO_DB’ environment variable to change the database directory. To use, set the ‘PATH_TO_DB’ environment variable to point to the database file, for example: PATH_TO_DB="./database/webhook.db". By default, the directory for the database file is “/server/database/webhook.db”.

If you would like to persist the database file between restarts of the Postee container, then you should use a persistent storage option to mount the "/server/database" directory of the container. The "deploy/kubernetes" directory in this project contains an example deployment that includes a basic Host Persistency.

Postee supports use of environment variables for Output fields: User, Password and Token. Add preffix $ to the environment variable name in the configuration file, for example:

outputs:
- name: my-jira   
  type: jira     
  enable: true
  user: $JIRA_USERNAME
  token: $JIRA_SERVER_TOKEN