Fuzzing for Security

Yesterday, I attended an internal workshop about ethical hacking. Hacking skills like fuzzing can be used to quantitatively assess and measure security threats in software.  Fuzzing is a software testing technique used to discover coding errors and security loopholes in software, operating systems or networks by injecting massive amounts of random data, called fuzz, to the system in an attempt to make it crash. If the program contains a vulnerability that can leads to an exception, crash or server error (in the case of web apps), it can be determined that a vulnerability has been discovered.

A fuzzer is a program that generates and injects random (and in general faulty) input to an application. Its main purpose is to make things easier and automated.

There are typically two methods for producing fuzz data that is sent to a target, Generation or Mutation. Generational fuzzers are capable of building the data being sent based on a data model provided by the fuzzer creator. Sometimes this is simple and dumb as sending random bytes, swapping bytes or much smarter by knowing good values and combining them in interesting ways.

Mutation on the other hand starts out with a known good “template” which is then modified. However, nothing that is not present in the “template” or “seed” will be produced.

Generally fuzzers are good at finding buffer overflow, DoS, SQL Injection, Format String bugs etc. They do a poor job at finding vulnerabilites related to information disclosure, encryption flaws and any other vulnerability that does not cause the program to crash.  Fuzzing is simple and offers a high benefit-to-cost ratio but does not replace other proven testing techniques.

What is your computer doing over the week-end ?

Using execution context ID (ECID)

Execution context ID (ECID) is a unique identifier to correlate events or requests associated with the same transation across several components.

The ECID value for a particular request is generated at the first layer and is passed down to the subsequent layers. The ECID value is logged (and auditable) in each product involved in the transaction. ECID allows an administrator to track the end-to-end flow of a particular request across the product stack.

ECID are supported by OUD and can be used to track LDAP requests from the client down to the ultimate LDAP server processing the request (inclusing LDAP access layer/proxy if any).

When performing a LDAP operation, a client can pass a ECID using the LDAP control extension 2.16.840.1.113894.1.8.31 . This ECID is logged by OUD. The OUD server generates a ECID in case none is present in the incoming request.

ECID are logged in the “Oracle Access Logger”. By default, this logger is disabled. To enable it, run the command below:

dsconfig set-log-publisher-prop \
         –publisher-name Oracle\ Access\ Logger \
         –set enabled:true\
–hostname localhost\
–port <admin port>\
–bindDN cn=Directory\ Manager \
–bindPassword ****** \

Here is a sniplet of the Oracle access log:

[2012-08-16T16:10:26.770+02:00] [OUD] [TRACE] [OUD-24641559] [PROTOCOL] [host: prehnite] [nwaddr:] [tid: 25] [userId: sduloutr] [ecid:,0] [category: REQ] [conn: -1] [op: 80] [msgID: 81] [dn: o=example] [type: synchronization] MODIFY

The administrator can then search the logs using a particular ECID value. Audit logs can be queried for a given ECID through Oracle BI Publisher’s audit reports. For example, if you send an LDAP request to Oracle Virtual Directory front-ending Oracle Unified Directory, an ECID associated with the LDAP request is present in the OVD diagnostic logs and audit logs; similarly, when the query reaches OUD, OUD includes the same ECID in its diagnostic logs and audit reports.

Monitoring OUD with VisualVM

VisualVM is a visual tool integrating several command line JDK tools and lightweight profiling capabilities. Designed for both production and development time use, it further enhances the capability of monitoring and performance analysis for the Java SE platform.

Here are the steps to use VisualVM to monitor Oracle Unified Directory:

#1 Download the latest release of VisualVM from http://visualvm.java.net/

#2 Enable the MBeans plugin as described in  http://visualvm.java.net/mbeans_tab.html to take advantage of the statistics exposed by OUD

#3 Enable JMX on OUD

  1. Start the server.
  2. Enable the JMX Connection Handler and set the port number to be used with JMX.

    Choose a port that is not in use and to which the user that is running the server has access rights.

    $ dsconfig -h localhost -p 4444 -D "cn=directory manager" -j pwd-file -X -n \
      set-connection-handler-prop --handler-name "JMX Connection Handler" \
      --set enabled:true --set listen-port:1689
  3. Add the JMX read, write, and notify privileges to the root DN.
    $ dsconfig -h localhost -p 4444 -D "cn=directory manager" -j pwd-file -X -n \
      set-root-dn-prop \
      --add default-root-privilege-name:jmx-read \
      --add default-root-privilege-name:jmx-write \
      --add default-root-privilege-name:jmx-notify
  4. Restart the server.

#4 Connect to OUD from VisualVM

To connect VisualVMto a server instance, click on File/Add JMX Connection. The following fields are required:

  • JMX URL:


    • host is a host name, an IPv4 numeric host address, or an IPv6 numeric address enclosed in square brackets.
    • port is the decimal port number of the JMX connector.

    The default JMX URL is:


  • User Name. A valid LDAP user name.

    The default Directory Manager user name is cn=Directory Manager.

  • Password. The user’s LDAP password.

#5 Browse MBeans attributes

Go to the right panel then select the MBean tab and navigate in the MBean tree. You can get access to the OUD config, monitoring information & statistics and all the convenient java metrics.
Note: You can plot those JMX numeric values in VisualVM that appear in bold. To do so, double-clicking on numeric attribute values will display a chart that plots changes in that numeric value.