Thursday, January 16, 2020

Two final 2019 CVEs for Apache CXF

Apache CXF 3.3.5 and 3.2.12 have been released. These releases contain fixes for two new security advisories:
  • CVE-2019-12423: Apache CXF OpenId Connect JWK Keys service returns private/secret credentials if configured with a jwk keystore. 
  • CVE-2019-17573: Apache CXF Reflected XSS in the services listing page. Note that this attack exploits a feature which is not typically not
    present in modern browsers, who remove dot segments before sending the
    request. However, Mobile applications may be vulnerable.
Please see the CXF security advisories page for information on all of the CVEs issued for Apache CXF over the years.

Tuesday, November 5, 2019

Two new CVEs released for Apache CXF

Apache CXF 3.3.4 and 3.2.11 have been released. Along with the usual bug fixes and dependency updates, these releases contain fixes for two new CVEs:
  • CVE-2019-12419: Apache CXF OpenId Connect token service does not properly validate the clientId. The problem here is that the OAuth access token service didn't validate that the submitted clientId matches that of the authenticated principal, thus allowing a malicious client to obtain an access token using a code issued to another client. Of course, this requires the malicious client to actually obtain the authorization code for the other client somehow.
  • CVE-2019-12406: Apache CXF does not restrict the number of message attachments. Essentially here CXF did not impose any restrictions on the number of message attachments, meaning that a malicious entity could try to attempt a denial of serice attack by generating a message with a huge number of message attachments.
Please update to the latest CXF releases to pick up fixes for these advisories.

Monday, August 26, 2019

Annotation support with Apache Shiro

Apache Shiro is a Java framework to simply authentication, authorization etc. I previously blogged about a test-case I wrote that shows how to use Shiro with Apache CXF to authenticate and authorize a username and password received as part of a web service request. This post extends the previous post by showing how to use Shiro to enable authorization via annotations on the service implementation.

The previous post defined some required roles for an endpoint in Spring, and passed them through to a ShiroUTValidator class which checks that the authenticated subject has all of the defined roles:

The problem with this approach is that it's not possible to specify individual roles for different methods in the service implementation - the user must have the role to invoke on any of the methods.

An alternative is instead to use Shiro's annotation support. Here we can add annotations to the service endpoint implementation to require that the authenticated user has the correct role (@RequiresRoles) or permissions (@RequiresPermissions). Note that these annotations are specific to Shiro, support is not yet added to support the standard annotations (see here).

So to change our test-case to use annotations, instead of defining the roles in Spring, we instead define the following annotation in the service implementation:

In the spring configuration for the service, we need to add a few additional interceptors so that the annotation gets processed:

That's all that's required to get Shiro annotations working with CXF service implementations. The full test source is available here.

Thursday, May 9, 2019

CoAP support in Apache Camel

The Constrained Application Protocol (CoAP) is standardized in RFC-7252. It offers REST-like functionality over UDP for constrained devices in the Internet of Things. Apache Camel has had support for the CoAP protocol since the 2.16 release, by using the eclipse Californium framework. It offers support for using CoAP in both producer and consumer mode, and also offers integration with the Camel REST DSL. In this post, we will cover a number of significant improvements to the Camel CoAP component for the forthcoming 3.0.0 release.

1) DTLS support

The first significant improvement is that the CoAP component has been updated to support DTLS, something that necessitated a major upgrade of the californium dependency. CoAP supports TLS / UDP using a "coaps" scheme, something it is now possible to use in Camel. To see how this all works, take a look at the following github test-case I put together:
  • camel-coap - A test-case for the camel-coap component. It shows how to use the coap component with the Camel REST DSL + TLS.
It follows the same approach as the previous tutorial I wrote on securing the Jetty component in Camel:

It uses the Camel REST DSL to create a simple REST service on the "/data" path that produces an XML response when invoked with a GET request. The actual route is omitted above, it just returns a XML document read in from a file. Note the component of the REST DSL is "coap" and it uses a scheme of "coaps". When using a scheme of "coaps", we need to configure the relevant TLS configuration, something that is done by referring to a "sslContextParameters" bean, which in this case contains a reference to a keystore used to retrieve the TLS key. Note that when using a certificate for TLS with CoAP, an elliptic curve key is required - RSA is not supported.

On the client side, the test-case shows how to configure a Camel producer to invoke on the coaps REST API:

Note that a scheme of "coaps" is used, and that it refers to a sslContextParameters bean containing the truststore to use, as well as a specific CipherSuite (this is optional, I just put it in to show how to use it). As the logging output does not really show what is going on, here is the Wireshark output which shows that DTLS is in use:

2) Support for Raw Public Keys and Pre-Shared Keys

In the section above, we saw how to configure CoAP with DTLS by referring to a sslContextParameters bean, which in turn refers to keystores to extract private keys and certificates. This is one option to support DTLS. However we also have two other options instead of using certificates.

The first is called Raw Public Keys. This is when we may not have access to a certificate containing the (trusted) public key of the endpoint. In this case, we can configure TLS using a PrivateKey and/or PublicKey objects. Both are required for a service. The client needs to be configured with a trustedRpkStore parameter, which is an interface supplied by Californium, that determines trust in an identifier. If the service is configured with "clientAuthentication" of "REQUIRE", then the service must configure trustedRpkStore, and the client must also specify a privateKey parameter. Here is a sample code snippet from the Camel tests:

The second option is called Pre-Shared Keys. This is when we don't have access either to certificates or public keys, but have some symmetric keys that are shared between both the client and service. In this case, we can use these keys for TLS. Both the client and service are configured with a "pskStore" parameter, which is an interface in Californium that associates a (byte[]) key with an identity. Here is a sample code snippet from the Camel tests:

3) Support for TCP / TLS

A newer RFC (RFC-8323) extends the original RFC to add support for CoAP over TCP and Websockets. Camel 3.0.0 has added support for using CoAP over both TCP and also TLS over TCP. Websocket support is not currently available. RFC-8323 uses two new schemes for TCP, both of which are supported in Camel - "coap+tcp" for CoAP / TCP, and "coaps+tcp" for CoAP / TCP with TLS.  Only the certificate method of configuring TLS is supported, which works in exactly the same way as for DTLS above. Pre-shared keys and Raw Public Keys are not supported over TCP now, only UDP.

To see how it works, simply alter the configuration in the github testcase and change "coaps" to "coaps+tcp" (in both locations). Now run the test-case again and it should work seemlessly:

Tuesday, April 30, 2019

Securing the Apache Camel REST DSL

Recently I put together a simple test-case for Apache Camel's REST DSL and realised that it illustrated quite a few security concepts, as well as various Camel components, that might be interesting to blog about. The test-case is a simple spring-based project, which is available on github here:
  • camel-jetty: A test-case for the Camel Jetty component, TLS, the REST DSL + Jasypt.
In particular, the Camel spring configuration is here. Let's take a look at the different pieces one by one.

1) The Apache Camel REST DSL

Apache Camel offers a REST DSL which makes it really easy to create a simple REST service.
Here we are creating a simple REST service on the "/data" path that produces an XML response when invoked with a GET request. The actual functionality is delegated to a Camel route called "direct:get":

Here we are reading in some files from a directory using the Camel File component and using "pollEnrich" to include the contents of that directory into the message that is returned to the user. Finally we need to tell Camel how to create the REST DSL. Camel supports a wide range of components, but for the purposes of this example we are using the Camel Jetty component:

Note the port is not hard-coded, but instead retrieved from a randomly generated property in the pom using the "reserve-network-port" goal of the "build-helper-maven-plugin".

2) Getting TLS to work with the Camel REST DSL

To support TLS with the Camel REST DSL, we need to set the scheme to "https" as above in the "restConfiguration". The REST configuration also refers to a property called "sslContextParameters", which is where we obtain the keys required to support TLS. See the Camel JSSE documentation for more information on this property.

The sslContextParameters bean definition allows us to define the key managers and trust managers for the TLS endpoint by referring to Java keystore files with the relevant passwords. If we are not supporting client authentication, the trustManagers portion can be omitted.

3) Using Jasypt to decrypt keystore passwords for use in TLS

Note above that we have not hard-coded the TLS keystore passwords in our Camel spring configuration, but are instead retrieving them from a property. Camel offers the ability to store the passwords in encrypted form, by using the Camel Jasypt component to decrypt them given a master password. The encrypted passwords themselves are stored in a file:

These encrypted passwords are obtained using the camel-jasypt jar (shipped for example in the Camel distribution):
  • java -jar camel-jasypt-2.23.1.jar -c encrypt -p master-secret -i storepass
To decrypt the passwords at runtime, we define the following bean in our Camel spring configuration:
This retrieves the master password from a system property. For the purposes of this demo, the password is set as a system property in the "maven-surefire-plugin" defined in the pom.

4) Invoking on our secured REST service using the Camel HTTP4 component

The demo also includes a client route which invokes on the secured REST service we have created. We use the Camel HTTP4 component for this:

We start the route using the Camel Timer component, before calling the HTTP4 component. As we have included a query String in the request URI, the http4 component will issue a GET request. As for the REST service, we need to configure the TLS keys using the "sslContextParameters" parameter.

On the client side we only need the trustManagers configuration, unless of course we want to support client authentication. For the purposes of this demo, we also need to configure a custom x509HostnameVerifier property. This is because the TLS certificate the service is using will not be accepted by the client by default, as the common name of the certificate does not match the domain name of the service. We can circumvent this  (for testing purposes only, it is not secure!) by using the following hostname verifier:

Finally we log the service response to the console so we can see the test-case working.

Friday, April 5, 2019

Performance gain for web service requests in Apache CXF

In this post I want to talk about a recent performance gain for JAX-WS web service requests I made in Apache CXF. It was prompted by a mail to the CXF users list. The scenario was for a JAX-WS web service where certain requests are secured using WS-SecurityPolicy, and other requests are not. The problem was that the user observed that the security interceptors were always invoked in CXF, even for the requests that had no security applied to the message, and that this resulted in a noticeable performance penalty for large requests.

The reason for the performance penalty is that CXF needs to convert the request into a Document Object Model to apply WS-Security (note there is also a streaming WS-Security implementation available, but the performance is roughly similar). CXF needs to perform this conversion as it requires access to the full Document to perform XML Signature verification, etc. on the request. So even for the insecure request, it would apply CXF's SAAJInInterceptor. Then it would iterate through the security headers of the request, find that there was none present, and skip security processing.

However when thinking about this problem, I realised that before invoking the SAAJInInterceptor, we could check to see whether a security header is actually present in the request (and whether it matches the configured "actor" if one is configured). CXF makes the message headers available in DOM form, but not the SOAP Body (unless SAAJInInterceptor is called). If no matching security header is available, then we can skip security processing, and instead just perform WS-SecurityPolicy assertion using a set of empty results.

This idea is implemented in CXF for the 3.3.2 release via the task CXF-8010. To test what happens, I added a test-case to github here. This creates a war file with a service with two operations, one that is not secured, and one that has a WS-SecurityPolicy asymmetric binding applied to the operations. Both operations contain two parameters, an integer and a String description.

To test it, I added a JMeter test-case here. It uses 10 threads to call the insecure operation 30,000 times. The description String in each request contains the URL encoded version of the WS-Security specification to test what happens with a somewhat large request.

Here are the results using CXF 3.3.1:
and here are the results using the CXF 3.3.2-SNAPSHOT code with the fix for CXF-8010 applied:
Using CXF 3.3.1 the throughput is 1604.25 requests per second, whereas with CXF 3.3.2 the throughput is 1795.26 requests per second, a gain of roughly 9%. For a more complex SOAP Body I would expect the gain to be a lot greater.

Friday, March 29, 2019

HTTP Signature support in Apache CXF

Apache CXF provides support for the HTTP Signatures draft spec since the 3.3.0 release. Up to this point, JAX-RS message payloads could be signed using either XML Security or else using JOSE. In particular, the JOSE functionality can be used to also sign HTTP headers. However it doesn't allow the possibility to sign the HTTP method and Path, something that HTTP Signature supports. In this post we'll look at how to use HTTP Signatures with Apache CXF.

I uploaded a sample project to github to see how HTTP Signature can be used with CXF:
  • cxf-jaxrs-httpsig: This project contains a test that shows how to use the HTTP Signature functionality in Apache CXF to sign a message to/from a JAX-RS service.

1) Client configuration

The client configuration to both sign the outbound request and verify the service response is configured in the test code:

Two JAX-RS providers are added - CreateSignatureInterceptor creates a signature on the outbound request, and VerifySignatureClientFilter verifies a signature on the response. The keys used to sign the request and verify the response are configured in properties files, that are referenced via the "" and "" configuration tags:

Here we can see that a keystore is being used to retrieve the private key for signing the outbound request. If you wish to retrieve keys from some other source, then instead of using configuration properties it's best to configure the MessageSigner class directly on the CreateSignatureInterceptor.

By default CXF will add all HTTP headers to the signature. In addition, a client will also include the HTTP method and path using the "(request-target)" header. Also if the payload is not empty, it will be digested with the digest added to a "Digest" HTTP header, which is also signed. This provides payload integrity. By default, the signature algorithm is "rsa-sha256", of course it is possible to configure this. A secured request using HTTP signature looks like the following:

2) Service configuration

The service configuration is defined in spring. Two different JAX-RS providers are used on the service side - VerifySignatureFilter is used to verify a signature on the client request, and CreateSignatureInterceptor is used to sign the response message as per the client request.

For more information on how to use HTTP Signatures with Apache CXF, refer to the CXF documentation.