Applies to: Microsoft Cloud App Security
Docker on Ubuntu and RHEL on-premises
OS: Ubuntu 14.04 or higher (there is no stable version of docker to support Ubuntu 17.10), or RHEL 7.2 or higher
Disk space: 250 GB
RAM: 4 GB
Set your firewall as described in Network requirements
Log collector performance
The Log collector can successfully handle log capacity of up to 50 GB per hour. The main bottlenecks in the log collection process are:
Network bandwidth - your network bandwidth determines the log upload speed.
I/O performance of the virtual machine allocated by your IT - determines the speed at which logs are written to the log collector’s disk. The log collector has a built-in safety mechanism that monitors the rate at which logs arrive and compares it to the upload rate. In cases of congestion, the log collector starts to drop log files. If your setup generally exceeds 50 GB per hour, it is recommended that you split the traffic between multiple log collectors.
Set up and configuration
Step 1 – Web portal configuration: Define data sources and link them to a log collector
Go to the automated upload setting page:
In the Cloud App Security portal, click the settings icon followed by Log collectors.
For each firewall or proxy from which you want to upload logs, create a matching data source:
a. Click Add data source.
b. Name your proxy or firewall.
c. Select the appliance from the Source list. If you select Custom log format to work with a network appliance that is not specifically listed, see Working with the custom log parser for configuration instructions.
d. Compare your log with the sample of the expected log format. If your log file format does not match this sample, you should add your data source as Other.
e. Set the Receiver type to either FTP, FTPS, Syslog – UDP, or Syslog – TCP, or Syslog – TLS.
Integrating with secure transfer protocols (FTPS and Syslog – TLS) often requires additional settings or your firewall/proxy.
f. Repeat this process for each firewall and proxy whose logs can be used to detect traffic on your network.
Go to the Log collectors tab at the top.
a. Click Add log collector.
b. Give the log collector a name.
c. Enter the Host IP address of the machine you will use to deploy the Docker.
The host IP address can be replaced with the machine name, if there is a DNS server (or equivalent) that will resolve the host name.
d. Select all Data sources that you want to connect to the collector, and click Update to save the configuration see the next deployment steps.
- A single Log collector can handle multiple data sources.
- Copy the contents of the screen because you will need the information when you configure the Log Collector to communicate with Cloud App Security. If you selected Syslog, this information will include information about which port the Syslog listener is listening on.
Further deployment information will appear. Copy the run command from the dialog. You can use the copy to clipboard icon .
Export the expected data source configuration. This configuration describes how you should set the log export in your appliances.
Step 2 – On-premises deployment of your machine
The following steps describe the deployment in Ubuntu. The deployment steps for other platforms are slightly different.
Open a terminal on your Ubuntu machine.
Change to root privileges using the command:
To bypass a proxy in your network, run the following two commands:
export http_proxy='<IP>:<PORT>' (e.g. 22.214.171.124:8888) export https_proxy='<IP>:<PORT>'
If you accept the software license terms, uninstall old versions and install Docker CE by running the following command:
curl -o /tmp/MCASInstallDocker.sh https://adaprodconsole.blob.core.windows.net/public-files/MCASInstallDocker.sh && chmod +x /tmp/MCASInstallDocker.sh; /tmp/MCASInstallDocker.sh
If this command fails to validate your proxy certificate, run the command using
curl -kat the beginning.
Deploy the collector image on the hosting machine by importing the collector configuration. Do this by copying the run command generated in the portal. If you need to configure a proxy add the proxy IP address and port number. For example, if your proxy details are 192.168.10.1:8080, your updated run command is:
(echo 6f19225ea69cf5f178139551986d3d797c92a5a43bef46469fcc997aec2ccc6f) | docker run --name MyLogCollector -p 21:21 -p 20000-20099:20000-20099 -e "PUBLICIP='126.96.36.199'" -e "PROXY=192.168.10.1:8080" -e "CONSOLE=tenant2.eu1-rs.adallom.com" -e "COLLECTOR=MyLogCollector" --security-opt apparmor:unconfined --cap-add=SYS_ADMIN --restart unless-stopped -a stdin -i microsoft/caslogcollector starter
Verify that the collector is running properly by running the following command:
docker logs \<collector_name\>
You should see the message: Finished successfully!
Step 3 - On-premises configuration of your network appliances
Configure your network firewalls and proxies to periodically export logs to the dedicated Syslog port of the FTP directory according to the directions in the dialog, for example:
BlueCoat_HQ - Destination path: \<<machine_name>>\BlueCoat_HQ\
Step 4 - Verify the successful deployment in the Cloud App Security portal
Check the collector status in the Log collector table and make sure the status is Connected. If it is Created, it is possible that the log collector connection and parsing has not completed.
You can also go to the Governance log and verify that logs are being periodically uploaded to the portal.
If you encounter problems during deployment, see Troubleshooting Cloud Discovery.
Optional - Create custom continuous reports
After you have verified that the logs are being uploaded to Cloud App Security and the reports are being generated, you can create custom reports. You can now create custom discovery reports based on Azure Active Directory user groups. For example, if you want to see the cloud use of your marketing department, you can import the marketing group using the import user group feature, and then create a custom report for this group. You can also customize a report based on IP address tag or IP address ranges.
- In the Cloud App Security portal, under the Settings cog, select Cloud Discovery settings and then select Manage continuous reports.
- Click the Create report button and fill in the fields.
- Under the Filters you can filter the data by data source, by imported user group, or by IP address tags and ranges.