Together with Splunk and TomTom we organized a Security Intelligence round table last week with a select number of large telecom providers, financial institutes and central government customers. The idea was to share experiences and listen to the latest security and compliance trends, hosted by our special guest Mark Seward, Director of Security and Compliance Marketing at Splunk.

During the first introductions the following topics of interest where mentioned:
- What techniques are available to correlate events from different sources?
- How can we monitor an outsourced environment?
- How can we setup a distributed security management solution?
- How can non-expert users use Splunk?
- What additional use cases are there?
- What can do Splunk to monitor the security of a VMware environment?
What is Security Intelligence?
Mark started off by saying that the Splunk language is basicly like a mix of unix and sql. It is a bit of Google meets MS Excel. The idea of Security Intelligence is that you start with high-level questions about what risk your business has and you should start thinking like a criminal when looking at these risks. Each industry has different risks. For instance:
- For an ISP the risk might be that a new user immediately setups multiple mail accounts indicating a spammer.
- In a Telco environment you want to find out if there is any abuse of service plans. Watch the CDR’s to look for people that call all over the world.
- In a banking environment you might want to look for ATM transactions from the same customer at different locations at the same time.
Today all data is security relevant. Security teams should work with all the other teams and companies should get rid of any boundaries. At some of the customers this is already happening: Every 6 months the team looks how hackers could get access to the banking application. This is now still a paper exercise. The latest trend is that development and security teams work together when building new apps so that the security team knows what to look for in the logging information.
Security intelligence is also about looking for unknown threats. So look more at DNS, DHCP, VPN proxy data (sources that users use every day) and less to firewall, IPS and antivirus (sources that look for known attacks).
Correlating with Splunk
Spunk has a very rich set of commands to answer these questions. There are 5 different correlation techniques where Mark has written a whitepaper about. One example is for instance nested searches to answer questions like who is using someone else’s credentials and what behavior am I seeing that looks like malware. Splunk has the idea to come up with a community where Splunk users can share searches. We all agree that this would be a great idea.
Another nice example is to find out who is accessing the servers without a VPN connection or physical access to the building. This means that someone jumped the gate or is using someone elses login credentials. Finding this out with Splunk means correlating VPN logs, active directory logs and the logs of the physical access. The different user data formats can be normalized via an external lookup.
It is also possible to anonymize data via the transform command. In this case certain users wil not be able to see user or credit card details. This data will be masked with *’s for these users.
Click here to download the Splunk datasheet for detecting Advanced Persistent Threats.
Splunk versus SIEM versus BI
A lot of questions can be answered with the traditional business intelligence (BI) tools. Normal BI tools need lot’s of time to implement and are hard to integrate. Splunk is like giving the security group a real-time BI tool. The advantage of using Splunk is that you have visibility to all the data. So for instance a spike in the traffic on the website could mean a lot of things. It could be an operational issue, the marketing department could run a promotion with generates a lot of extra customer traffic or it could be a security issue. The power of Splunk is that different teams can look at the data in realtime.
The Splunk app for enterprise security was also briefly discussed. This is one of the few payed Splunk apps (most are free) that provide SIEM like capabilities to the core Splunk platform:
- This app normalizes all the sources for known threats like firewall, IDS etc. The standard Splunk functionality can already be used for looking at the unknown threats;
- It supports incident handling;
- It has functions to add urgency / criticality to servers and events;
- It contains more than 50 predifined reports and dashboards;
- It has additional correlation and workflow actions.
Click here for more information about the Splunk app for Enterprise Security
During the discussion the idea came up to install the Splunk app for Enterprise Security in the SMT lab so that customers can bring their data and see how this app would work in their environment. So watch out for more information about SMT’s Splunk CSI lab.
The result
At the end of this meeting the following (action) points were taken home by the group:
- Setup hacking sessions with multiple teams
- Want to get more data sources into Splunk
- Get also data out of the cloud into Splunk
- A lot more use cases for Splunk
A great session, with special thanks to the guests, to Mark Seward from Splunk for coming over all the way from the US and to TomTom for hosting the session!