Playbook 1: SQS Public Policy

Introduction

This playbook describes how to configure Dow Jones Hammer to identify SQS queues that are accessible because of the policy settings.

1. Issue Identification

Dow Jones Hammer investigates policy statements for SQS queues and checks whether both of the following conditions apply:

1. The Principal parameter value is * or {"AWS": "*"} (these are identical in terms of the issue’s definition)
2. Statement is not restricted by IpAddress condition or IpAddress condition is set to {"aws:SourceIp": "0.0.0.0/0"}.

When Dow Jones Hammer detects an issue, it writes the issue to the designated DynamoDB table.

According to the Dow Jones Hammer architecture, the issue identification functionality uses two Lambda functions. The table lists the Python modules that implement this functionality:

2. Issue Reporting

You can configure automatic reporting of cases when Dow Jones Hammer identifies an issue of this type. Dow Jones Hammer supports integration with JIRA and Slack. These types of reporting are independent from one another and you can turn them on/off in the Dow Jones Hammer configuration.

Thus, in case you have turned on the reporting functionality for this issue and configured corresponding integrations, Dow Jones Hammer, as defined in the configuration, can:

• raise a JIRA ticket and assign it to a specific person in your organization;
• send the issue notification to the Slack channel or directly to a Slack user.

Additionally Dow Jones Hammer tries to detect person to report issue to by examining owner tag on affected SQS queue. In case when such tag exists and is valid JIRA/Slack user:

• for JIRA: jira_owner parameter from ticket_owners.json is ignored and discovered owner is used instead as a JIRA assignee;
• for Slack: discovered owner is used in addition to slack_owner value from ticket_owners.json.

This Python module implements the issue reporting functionality:

hammer/reporting-remediation/reporting/create_sqs_policy_issue_tickets.py

3. Issue Remediation

3.1 Automatic

To reduce the workload of your DevOps engineers and mitigate the threats stemming from this issue, you can configure automatic remediation of issues. It means that in case Dow Jones Hammer has detected and reported an issue, but the assignee of the report has not remediated the issue within a timeframe specified in the configuration, the Dow Jones Hammer remediation job will adjust SQS queue policy to eliminate this vulnerability.

In this specific case, Dow Jones Hammer restricts public statement by adding (or changing) IpAddress condition value that allow access only for IP addresses defined in RFC 1918 - Address Allocation for Private Internets.

Dow Jones Hammer will save the pre-remediation SQS queue policy configuration to a S3 backup bucket. You can use this configuration to rollback the automatic remediation manually, if necessary.

This Python module implements the issue remediation functionality:

hammer/reporting-remediation/remediation/clean_sqs_policy_permissions.py

3.2 Manual

To retain full control on the remediation functionality you can disable automatic remediation in config.json and launch it manually:

1. Login to Dow Jones Hammer reporting and remediation EC2 via SSH with centos user and ssh key you created during deployment: ssh -l centos -i <private_key> <EC2_IP_Address>
2. Become root user: sudo su -
3. Change directory to Dow Jones Hammer sources: cd /hammer-correlation-engine
4. Launch Dow Jones Hammer remediation script: python3.6 -m remediation.clean_sqs_policy_permissions
5. Confirm or refuse remediation of each issue separately

4. Setup Instructions For This Issue

To configure the detection, reporting, and remediation of this issue, you should edit the following sections of the Dow Jones Hammer configuration files:

4.1. The config.json File

The config.json file is the main configuration file for Dow Jones Hammer that is available at deployment/terraform/accounts/sample/config/config.json. To identify, report, and remediate issues of this type, you should add the following parameters in the sqs_queue_policy section of the config.json file:

Sample config.json section:

"sqs_public_access": {
        "enabled": false,
        "ddb.table_name": "hammer-sqs-public-access",
        "reporting": false,
        "remediation": false,
        "remediation_retention_period": 0
    },

4.2. The whitelist.json File

You can define exceptions to the general automatic remediation settings for specific SQS queues. To configure such exceptions, you should edit the sqs_policy section of the whitelist.json configuration file as follows:

Sample whitelist.json section:

"sqs_policy": {
    "123456789012": ["queue_url1", "queue_url1"]
}

4.3. The ticket_owners.json File

You should use the ticket_owners.json file to configure the integration of Dow Jones Hammer with JIRA and/or Slack for the issue reporting purposes.

You can configure these parameters for specific AWS accounts and globally. Account-specific settings precede the global settings in the ticket_owners.json configuration file.

Check the following table for parameters:

Sample ticket_owners.json section:

Account-specific settings:

{
    "account": {
        "123456789012": {
            "jira_project": "",
            "jira_owner": "Support-Cloud",
            "jira_parent_ticket": "",
            "slack_owner": ""
        }
    },
    "jira_project": "AWSSEC",
    "jira_owner": "Support-General",
    "jira_parent_ticket": "AWSSEC-1234",
    "slack_owner": ["#devops-channel", "bob"]
}

5. Logging

Dow Jones Hammer uses CloudWatch Logs for logging purposes.

Dow Jones Hammer automatically sets up CloudWatch Log Groups and Log Streams for this issue when you deploy Dow Jones Hammer.

5.1. Issue Identification Logging

Dow Jones Hammer issue identification functionality uses two Lambda functions:

• Initialization: this Lambda function selects slave accounts to check for this issue as designated in the Dow Jones Hammer configuration files and triggers the check.
• Identification: this Lambda function identifies this issue for each account/region selected at the previous step.

You can see the logs for each of these Lambda functions in the following Log Groups:

5.2. Issue Reporting/Remediation Logging

Dow Jones Hammer issue reporting/remediation functionality uses /aws/ec2/hammer-reporting-remediation CloudWatch Log Group for logging. The Log Group contains issue-specific Log Streams named as follows:

5.3. Slack Reports

In case you have enabled Dow Jones Hammer and Slack integration, Dow Jones Hammer sends notifications about issue identification and remediation to the designated Slack channel and/or recipient(s).

Check ticket_owners.json configuration for further guidance.

5.4. Using CloudWatch Logs for Dow Jones Hammer

To access Dow Jones Hammer logs, proceed as follows:

1. Open AWS Management Console.
2. Select CloudWatch service.
3. Select Logs from the CloudWatch sidebar.
4. Select the log group you want to explore. The log group will open.
5. Select the log stream you want to explore.

Check CloudWatch Logs documentation for further guidance.

6. Issue specific details in DynamoDB

Dow Jones Hammer stores various issue specific details in DynamoDB as a map under issue_details key. You can use it to create your own reporting modules.