Colm O hEigeartaigh

Securing Apache Hadoop Distributed File System (HDFS) - part IV

Colm O hEigeartaigh - Wed, 04/26/2017 - 18:39
This is the fourth in a series of blog posts on securing HDFS. The first post described how to install Apache Hadoop, and how to use POSIX permissions and ACLs to restrict access to data stored in HDFS. The second post looked at how to use Apache Ranger to authorize access to data stored in HDFS. The third post looked at how Apache Ranger can create "tag" based authorization policies for HDFS using Apache Atlas. In this post I will look at how you can implement transparent encryption in HDFS using the Apache Ranger Key Management Service (KMS).

1) Install and Configure the Apache Ranger KMS

If you have not done so already, then follow the instructions in this tutorial to install the Apache Ranger admin service, and then start it via "sudo ranger-admin start". Open a browser and go to "http://localhost:6080/". Log on with "admin/admin" and click on "Settings". Create a new user corresponding to the name of the user which starts HDFS.

The next step is to install the Apache Ranger KMS. Please follow step (2) in a blog post I wrote last year about this. When installation is complete, then start the KMS service with "sudo ranger-kms start". Log out of the Admin UI and then log back in again with the credentials "keyadmin/keyadmin". Click on the "+" button on the "KMS" tab to create a new KMS Service. Specify the following values:
  • Service Name: kmsdev
  • KMS URL: kms://http@localhost:9292/kms
  • Username: keyadmin
  • Password: keyadmin
When the "kmsdev" service has been created then click on it and edit the default policy that has been created. Edit the existing "allow condition" for "hdfs" adding in the user that will be starting HDFS (if not the "hdfs" user itself). Also grant the "CREATE" permission to that user so that we can create keys from the command line, and the "DECRYPT EEK" permission, so that the user can decrypt the data encryption key:


2) Create an encryption zone in HDFS

In your Hadoop distribution (after first following the steps in the first post), edit 'etc/hadoop/core-site.xml' and add the following property:
  • hadoop.security.key.provider.path - kms://http@localhost:9292/kms
Similarly, edit 'etc/hadoop/hdfs-site.xml' and add the following property:
  • dfs.encryption.key.provider.uri - kms://http@localhost:9292/kms
Start HDFS via 'sbin/start-dfs.sh'. Let's create a new encryption key called "enckey" as follows:
  • bin/hadoop key create enckey
If you go back to the Ranger Admin UI and click on "Encryption / Key Manager" and select the "kmsdev" service, you should be able to see the new key that was created. Now let's create a new encryption zone in HDFS as follows:
  • bin/hadoop fs -mkdir /zone
  • bin/hdfs crypto -createZone -keyName enckey -path /zone
  • bin/hdfs crypto -listZones
That's it! We can put data into the '/zone' directory and it will be encrypted by a key which in turn is encrypted by the key we have created and stored in the Ranger KMS.
Categories: Colm O hEigeartaigh

Securing Apache Hadoop Distributed File System (HDFS) - part III

Colm O hEigeartaigh - Fri, 04/21/2017 - 11:54
This is the third in a series of posts on securing HDFS. The first post described how to install Apache Hadoop, and how to use POSIX permissions and ACLs to restrict access to data stored in HDFS. The second post looked at how to use Apache Ranger to authorize access to data stored in HDFS. In this post we will look at how Apache Ranger can create "tag" based authorization policies for HDFS using Apache Atlas. For information on how to create tag-based authorization policies for Apache Kafka, see a post I wrote earlier this year.

The Apache Ranger admin console allows you to create security policies for HDFS by associating a user/group with some permissions (read/write/execute) and a resource, such as a directory or file. This is called a "Resource based policy" in Apache Ranger. An alternative is to use a "Tag based policy", which instead associates the user/group + permissions with a "tag". You can create and manage tags in Apache Atlas, and Apache Ranger supports the ability to imports tags from Apache Atlas via a tagsync service, something we will cover in this post.

1) Start Apache Atlas and create entities/tags for HDFS

First let's look at setting up Apache Atlas. Download the latest released version (0.8-incubating) and extract it. Build the distribution that contains an embedded HBase and Solr instance via:
  • mvn clean package -Pdist,embedded-hbase-solr -DskipTests
The distribution will then be available in 'distro/target/apache-atlas-0.8-incubating-bin'. To launch Atlas, we need to set some variables to tell it to use the local HBase and Solr instances:
  • export MANAGE_LOCAL_HBASE=true
  • export MANAGE_LOCAL_SOLR=true
Now let's start Apache Atlas with 'bin/atlas_start.py'. Open a browser and go to 'http://localhost:21000/', logging on with credentials 'admin/admin'. Click on "TAGS" and create a new tag called "Data".  Click on "Search" and the "Create new entity" link. Select an entity type of "hdfs_path" with the following values:
  • QualifiedName: data@cl1
  • Name: Data
  • Path: /data
Once the new entity has been created, then click on "+" beside "Tags" and associate the new entity with the "Data" tag.

2) Use the Apache Ranger TagSync service to import tags from Atlas into Ranger

To create tag based policies in Apache Ranger, we have to import the entity + tag we have created in Apache Atlas into Ranger via the Ranger TagSync service. First, start the Apache Ranger admin service and rename the HDFS service we created in the previous tutorial from "HDFSTest" to "cl1_hadoop". This is because the Tagsync service will sync tags into the Ranger service that corresponds to the suffix of the qualified name of the tag with "_hadoop". Also edit 'etc/hadoop/ranger-hdfs-security.xml' in your Hadoop distribution and change the "ranger.plugin.hdfs.service.name" to "cl1_hadoop". Also change the "ranger.plugin.hdfs.policy.cache.dir" along the same lines. Finally, make sure the directory '/etc/ranger/cl1_hadoop/policycache' exists and the user you are running Hadoop as can write and read from this directory.

After building Apache Ranger then extract the file called "target/ranger-<version>-tagsync.tar.gz". Edit 'install.properties' as follows:
  • Set TAG_SOURCE_ATLAS_ENABLED to "false"
  • Set TAG_SOURCE_ATLASREST_ENABLED to  "true"
  • Set TAG_SOURCE_ATLASREST_DOWNLOAD_INTERVAL_IN_MILLIS to "60000" (just for testing purposes)
  • Specify "admin" for both TAG_SOURCE_ATLASREST_USERNAME and TAG_SOURCE_ATLASREST_PASSWORD
Save 'install.properties' and install the tagsync service via "sudo ./setup.sh". It can now be started via "sudo ranger-tagsync-services.sh start".

3) Create Tag-based authorization policies in Apache Ranger

Now let's create a tag-based authorization policy in the Apache Ranger admin UI. Click on "Access Manager" and then "Tag based policies". Create a new Tag service called "HDFSTagService". Create a new policy for this service called "DataPolicy". In the "TAG" field enter a capital "D" and the "Data" tag should pop up, meaning that it was successfully synced in from Apache Atlas. Create an "Allow" condition for the user "bob" with component permission of "HDFS" and "read" and "execute":


The last thing we need to do is to go back to the Resource based policies and edit "cl1_hadoop" and select the tag service we have created above.

4) Testing authorization in HDFS using our tag based policy

Wait until the Ranger authorization plugin syncs the new authorization policies from the Ranger Admin service and then we can test authorization. In the previous tutorial we showed that the file owner and user "alice" can read the data stored in '/data', but "bob" could not. Now we should be able to successfully read the data as "bob" due to the tag based authorization policy we have created:
  • sudo -u bob bin/hadoop fs -cat /data/LICENSE.txt
Categories: Colm O hEigeartaigh

Securing Apache Hadoop Distributed File System (HDFS) - part II

Colm O hEigeartaigh - Thu, 04/20/2017 - 16:23
This is the second in a series of posts on securing HDFS. The first post described how to install Apache Hadoop, and how to use POSIX permissions and ACLs to restrict access to data stored in HDFS. In this post we will look at how to use Apache Ranger to authorize access to data stored in HDFS. The Apache Ranger Admin console allows you to create policies which are retrieved and enforced by a HDFS authorization plugin. Apache Ranger allows us to create centralized authorization policies for HDFS, as well as an authorization audit trail stored in SOLR or HDFS.

1) Install the Apache Ranger HDFS plugin

First we will install the Apache Ranger HDFS plugin. Follow the steps in the previous tutorial to setup Apache Hadoop, if you have not done this already. Then download Apache Ranger and verify that the signature is valid and that the message digests match. Due to some bugs that were fixed for the installation process, I am using version 1.0.0-SNAPSHOT in this post. Now extract and build the source, and copy the resulting plugin to a location where you will configure and install it:
  • mvn clean package assembly:assembly -DskipTests
  • tar zxvf target/ranger-1.0.0-SNAPSHOT-hdfs-plugin.tar.gz
  • mv ranger-1.0.0-SNAPSHOT-hdfs-plugin.tar.gz ${ranger.hdfs.home}
Now go to ${ranger.hdfs.home} and edit "install.properties". You need to specify the following properties:
  • POLICY_MGR_URL: Set this to "http://localhost:6080"
  • REPOSITORY_NAME: Set this to "HDFSTest".
  • COMPONENT_INSTALL_DIR_NAME: The location of your Apache Hadoop installation
Save "install.properties" and install the plugin as root via "sudo ./enable-hdfs-plugin.sh". The Apache Ranger HDFS plugin should now be successfully installed. Start HDFS with:
  • sbin/start-dfs.sh
2) Create authorization policies in the Apache Ranger Admin console

Next we will use the Apache Ranger admin console to create authorization policies for our data in HDFS. Follow the steps in this tutorial to install the Apache Ranger admin service. Start the Apache Ranger admin service with "sudo ranger-admin start" and open a browser and go to "http://localhost:6080/" and log on with "admin/admin". Add a new HDFS service with the following configuration values:
  • Service Name: HDFSTest
  • Username: admin
  • Password: admin
  • Namenode URL: hdfs://localhost:9000
Click on "Test Connection" to verify that we can connect successfully to HDFS + then save the new service. Now click on the "HDFSTest" service that we have created. Add a new policy for the "/data" resource path for the user "alice" (create this user if you have not done so already under "Settings, Users/Groups"), with permissions of "read" and "execute".


3) Testing authorization in HDFS

Now let's test the Ranger authorization policy we created above in action. Note that by default the HDFS authorization plugin checks for a Ranger authorization policy that grants access first, and if this fails it falls back to the default POSIX permissions. The Ranger authorization plugin will pull policies from the Admin service every 30 seconds by default. For the "HDFSTest" example above, they are stored in "/etc/ranger/HDFSTest/policycache/" by default. Make sure that the user you are running Hadoop as can access this directory.

Now let's test to see if I can read the data file as follows:
  • bin/hadoop fs -cat /data/LICENSE* (this should work via the underlying POSIX permissions)
  • sudo -u alice bin/hadoop fs -cat /data/LICENSE* (this should work via the Ranger authorization policy)
  • sudo -u bob bin/hadoop fs -cat /data/LICENSE* (this should fail as we don't have an authorization policy for "bob").

Categories: Colm O hEigeartaigh

Securing Apache Hadoop Distributed File System (HDFS) - part I

Colm O hEigeartaigh - Wed, 04/19/2017 - 17:49
Last year, I wrote a series of articles on securing Apache Kafka using Apache Ranger and Apache Sentry. In this series of posts I will look at how to secure the Apache Hadoop Distributed File System (HDFS) using Ranger and Sentry, such that only authorized users can access data stored in it. In this post we will look at a very basic way of installing Apache Hadoop and accessing some data stored in HDFS. Then we will look at how to authorize access to the data stored in HDFS using POSIX permissions and ACLs.

1) Installing Apache Hadoop

The first step is to download and extract Apache Hadoop. This tutorial uses version 2.7.3. The next step is to configure Apache Hadoop as a single node cluster so that we can easily get it up and running on a local machine. You will need to follow the steps outlined in the previous link to install ssh + pdsh. If you can't log in to localhost without a password ("ssh localhost") then you need to follow the instructions given in the link about setting up passphraseless ssh.

In addition, we want to run Apache Hadoop in pseudo-distributed mode, where each Hadoop daemon runs as a separate Java process. Edit 'etc/hadoop/core-site.xml' and add:
Next edit 'etc/hadoop/hdfs-site.xml' and add:

Make sure that the JAVA_HOME variable in 'etc/hadoop/hadoop-env.sh' is correct, and then format the filesystem and start Hadoop via:
  • bin/hdfs namenode -format
  • sbin/start-dfs.sh
To confirm that everything is working correctly, you can open "http://localhost:50090" and check on the status of the cluster there. Once Hadoop has started then upload and then access some data to HDFS:
  • bin/hadoop fs -mkdir /data
  • bin/hadoop fs -put LICENSE.txt /data
  • bin/hadoop fs -ls /data
  • bin/hadoop fs -cat /data/*
2) Securing HDFS using POSIX Permissions

We've seen how to access some data stored in HDFS via the command line. Now how can we create some authorization policies to restrict how to access this data? Well the simplest way is to use the standard POSIX Permissions. If we look at the /data directory we see that it has the following permissions "-rw-r--r--", which means other users can read the LICENSE file stored there. Remove access to other users apart from the owner via:
  • bin/hadoop fs -chmod og-r /data
Now create a test user called "alice" on your system and try to access the LICENSE we uploaded above via:
  • sudo -u alice bin/hadoop fs -cat /data/*
You will see an error that says "cat: Permission denied: user=alice, access=READ_EXECUTE".

3) Securing HDFS using ACLs

Securing access to data stored in HDFS via POSIX permissions works fine, however it does not allow you for example to specify fine-grained permissions for users other than the file owner. What if we want to allow "alice" from the previous section to read the file but not "bob"? We can achieve this via Hadoop ACLs. To enable ACLs, we will need to add a property called "dfs.namenode.acls.enabled" with value "true" to 'etc/hadoop/hdfs-site.xml' + re-start HDFS.

We can grant read access to 'alice' via:
  • bin/hadoop fs -setfacl -m user:alice:r-- /data/*
  • bin/hadoop fs -setfacl -m user:alice:r-x /data
To check to see the new ACLs associated with LICENSE.txt do:
  • bin/hadoop fs -getfacl /data/LICENSE.txt
In addition to the owner, we now have the ACL "user:alice:r--". Now we can read the data as "alice". However another user "bob" cannot read the data. To avoid confusion with future blog posts on securing HDFS, we will now remove the ACLs we added via:
  • bin/hadoop fs -setfacl -b /data
  • bin/hadoop fs -setfacl -b /data/LICENSE.txt
Categories: Colm O hEigeartaigh

Apache CXF 3.1.11 released

Colm O hEigeartaigh - Tue, 04/18/2017 - 14:11
Apache CXF 3.1.11 (and 3.0.13) has been released. This release fixes a large number of bugs (there are over a 100 issues fixed in the CXF JIRA for this release). From a security POV, here are some of the more notable bug fixes and changes:
  • CXF-7315 - Abstract the STS client token caching behaviour to allow the user to plug in a custom implementation
  • CXF-7296 - Add support to enable revocation for TLS via configuration (see here). 
  • CXF-7314 - Custom BinarySecurityTokens are not used to set up the security context
  • CXF-4692 - Allow customization of Request Security Token Response
  • CXF-7252 - TLSParameterJaxBUtils.getTrustManagers getting password from wrong system property
In addition, two new security advisories have been issued for bugs fixed in this release:
  • CVE-2017-5653 - Apache CXF JAX-RS XML Security streaming clients do not validate that the service response was signed or encrypted.
  • CVE-2017-5656 - Apache CXF's STSClient uses a flawed way of caching tokens that are associated with delegation tokens.
Please update to the latest releases if you are affected by either of these issues.
Categories: Colm O hEigeartaigh

Using OCSP with TLS in Apache CXF

Colm O hEigeartaigh - Thu, 03/30/2017 - 13:35
The previous article showed how to enable OCSP for WS-Security based SOAP services in Apache CXF, by checking the revocation status of a certificate used for X.509 digital signature. The article stated that OCSP is supported in Apache CXF when TLS is used to secure communication between a web service client and server, but didn't give any further information. In this post we will show how to enable OCSP when using TLS for both a web service (JAX-WS or JAX-RS) client and server.

The test-code is available on github here (also contains WS-Security OCSP tests):
  • cxf-ocsp: This project contains a number of tests that show how a CXF service can validate client certificates using OCSP.
1) Enabling OCSP for web service clients

First we'll look at enabling OCSP for web service clients. The TLSOCSPTest shows how this can be done. Two Java security properties are set in the test-code to enable OCSP: 
  • "ocsp.responderURL": The URL of the OCSP service
  • "ocsp.enable": "true" to enable OCSP
The first property is required if the service certificate does not contain the URL of the OCSP service in a certificate extension. Before running the test, install openssl and run the following command from the "openssl" directory included in the project (use the passphrase "security"):
  • openssl ocsp -index ca.db.index -port 12345 -text -rkey wss40CAKey.pem -CA wss40CA.pem -rsigner wss40CA.pem
Two options are available to get OCSP working for a web service client. The first is to configure TLS in code as shown in the first test contained in TLSOCSPTest. A PKIXBuilderParameters instance is created with the truststore and revocation is explicitly "enabled" on it. This is then wrapped in a CertPathTrustManagerParameters and used to initialise the TrustManagerFactory. 

The second test shows a new and alternative way of enabling OCSP if you want to configure your TLS keys in spring. This feature is only available from CXF 3.1.11 onwards.  The spring configuration file for the client contains a tlsClientParameters Element with the attribute "enableRevocation="true"". Once the "ocsp.enable" security property is set, then this will enable revocation checking on the certificate presented by the server during the TLS handshake.

2) Enabling OCSP for web service servers

We also show via the TLSOCSPClientAuthTest how to enable OCSP for web service servers that use CXF's Jetty transport. Openssl should be started as per the client tests. The server requires client authentication and then uses OCSP to verify the revocation status of the certificate presented by the client during the TLS handshake. The TLS configuration for the server is done in code. However it can also be done in spring using the "enableRevocation" attribute as per the client above.
Categories: Colm O hEigeartaigh

Using OCSP with WS-Security in Apache CXF

Colm O hEigeartaigh - Tue, 03/21/2017 - 16:32
The OCSP (Online Certificate Status Protocol) is a http-based protocol to check whether a given X.509 certificate is revoked or not. It is supported in Apache CXF when TLS is used to secure communication between a web service client and server. However, it is also possible to use with a SOAP request secured with WS-Security. When the client signs a portion of the SOAP request using XML digital signature, then the service can be configured to check whether the certificate in question is revoked or not via OCSP. We will cover some simple test-cases in this post that show how this can be done.

The test-code is available on github here:
  • cxf-ocsp: This project contains a number of tests that show how a CXF service can validate client certificates using OCSP.
The project contains two separate test-classes for WS-Security in particular. Both are for a simple "double it" SOAP web service invocation using Apache CXF. The clients are configured with CXF's WSS4JOutInterceptor, to encrypt and sign the SOAP Body using credentials contained in keystores. For signature, the signing certificate is included in the security header of the request. On the receiving side, the services are configured to validate the signature and to decrypt the request. In particular, the property "enableRevocation" is set to "true" to enable revocation checking.

The first test, WSSecurityOCSPTest, is a conventional test of the OCSP functionality. Two Java security properties are set in the test-code to enable OCSP (the server runs in the same process as the client):
  • "ocsp.responderURL": The URL of the OCSP service
  • "ocsp.enable": "true" to enable OCSP
The first property is required if the client certificate does not contain the URL of the OCSP service in a certificate extension. Before running the test, install openssl and run the following command from the "openssl" directory included in the project (use the passphrase "security"):
  • openssl ocsp -index ca.db.index -port 12345 -text -rkey wss40CAKey.pem -CA wss40CA.pem -rsigner wss40CA.pem
Now run the test (e.g.  mvn test -Dtest=WSSecurityOCSPTest). In the openssl console window you should see the OCSP request data.

The second test, WSSecurityOCSPCertTest, tests the scenario where the OCSP service signs the response with a different certificate to that of the issuer of the client certificate. Under ordinary circumstances, OCSP revocation checking will fail, and indeed this is tested in the test above. However it's also possible to support this scenario, by adding the OCSP certificate to the service truststore (this is already done in the test), and to set the following additional security properties:
  • "ocsp.responderCertIssuerName": DN of the issuer of the cert
  • "ocsp.responderCertSerialNumber": Serial number of the cert
Launch Openssl from the "openssl" directory included in the project:
  • openssl ocsp -index ca.db.index -port 12345 -text -rkey wss40key.pem -CA wss40CA.pem -rsigner wss40.pem
and run the test via "mvn test -Dtest=WSSecurityOCSPCertTest".
Categories: Colm O hEigeartaigh

WS-Security with MTOM support in Apache CXF 3.2.0

Colm O hEigeartaigh - Tue, 02/21/2017 - 16:43
Getting WS-Security to work with MTOM-enabled web services has been a long-standing feature request in Apache CXF. A couple of years ago, support was added to CXF and WSS4J to store raw cipher data in message attachments when MTOM is enabled, to avoid the cost of BASE-64 encoding the bytes and inlining them in the message. However, CXF did not support signing/encrypting content that contained xop:Include Elements (properly). In this case, just the references were signed/encrypted and not the attachments themselves (the user was alerted to this via a warning log). From Apache CXF 3.2.0, WS-Security with MTOM will be properly supported, something we will cover in this post.

1) Securing an MTOM-enabled message with WS-Security

Let's look at the outbound case first. There is a new configuration option in WSS4J 2.2.0:
  • expandXOPInclude: Whether to search for and expand xop:Include Elements for encryption and signature (on the outbound side). This means that the referenced bytes are encrypted/signed, and not just the references. The default is false on the outbound side in WSS4J.
CXF will set this configuration option to "true" automatically for both the "action" based and WS-SecurityPolicy based approaches if MTOM is enabled. Note that this configuration option also applies on the inbound side with slightly different semantics (see below).

The way this configuration option works is that it scans all children of all message elements to be signed/encrypted, and inlines any xop:Include bytes that it finds before signature/encryption. For the encryption case, if the "storeBytesInAttachment" configuration option is set to true (false in WSS4J, true by default in CXF if MTOM is enabled), the encrypted bytes are then stored in a message attachment. For signature, the original Element is retained and the inlined version is discarded and not included in the request, meaning that the signed bytes are not modified as a message attachment.

2) Validating an MTOM-enabled message with WS-Security

On the inbound side, the "expandXOPInclude" configuration option also applies:
  • expandXOPInclude: Whether to search for and expand xop:Include Elements prior in signed elements to signature verification. The default is "true". Note that this replaces the previous "expandXOPIncludeForSignature" configuration option prior to WSS4J 2.2.0.
CXF overrides this default behaviour by only setting "expandXOPInclude" to "true" on the inbound side if MTOM is enabled. So to summarize, if you wish to support WS-Security with MTOM in CXF from the (future) 3.2.0 release, you don't need to set any configuration option by default to get it to properly sign and encrypt the message bytes. CXF will take care of setting everything up for you.
Categories: Colm O hEigeartaigh

Securing an Apache Kafka broker using Apache Ranger and Apache Atlas

Colm O hEigeartaigh - Fri, 02/17/2017 - 17:28
Last year, I wrote a series of articles on securing Apache Kafka. In particular, the third article looked at how to use Apache Ranger to create authorization policies for Apache Kafka in the Ranger security admin UI, and how to install the Ranger plugin for Kafka so that it picks up and enforces the authorization policies. In this article, we will cover an alternative way of creating and enforcing authorization policies in Apache Ranger for Apache Kafka using Apache Atlas.

The Apache Ranger security admin UI allows you to assign users or groups a particular permission associated with a given Kafka topic. This is what is called a "Resource Based Policy" in Apache Ranger. However an alternative is also available called a "Tag Based Policy". Instead of explicitly associating the user/group + permission with a resource (such as a Kafka topic), instead we can associate the user/group + permission with a "tag" (we can also create "deny" based policies associated with a "tag"). The "tag" itself contains the information about the resource that is being secured.

How does Apache Ranger obtain the relevant tags and associated entities? This is where Apache Atlas comes in. The previous post described how to secure access to Apache Atlas using Apache Ranger. Apache Atlas allows you to associate "tags" with entities such as Kafka topics, Hive tables, etc. Apache Ranger provides a "tagsync" service which runs periodically and obtains the tags from Apache Atlas and uploads them to Apache Ranger. The Ranger authorization plugin for Kafka downloads the authorization policies, including tags, from the Ranger admin service and evaluates whether access is allowed or not based on the policy evaluation. Let's look at an example...

1) Start Apache Atlas and create entities/tags for Kafka

The first step is to start Apache Atlas as per the previous tutorial. Note that we are not using the Apache Ranger authorization plugin for Atlas, so there is no need to follow step 2). Next we need to upload the Kafka entity of type "kafka_topic" that we are interested in securing. That can be done via the following command:
  • curl -v -H 'Accept: application/json, text/plain, */*' -H 'Content-Type: application/json;  charset=UTF-8' -u admin:admin -d @kafka-create.json http://localhost:21000/api/atlas/entities
where "kafka-create.json" is defined as:
Once this is done, log in to the admin console using credentials "admin/admin" at http://localhost:21000. Click on "Tags" and "Create Tag" called "KafkaTag". Next go to "Search" and search for the entity we have uploaded ("KafkaTest"). Click on the "+" button under "Tags" and associate the entity with the tag we have created.


2) Start Apache Ranger and create resource-based authorization policies for Kafka

Next we will follow the first tutorial to install Apache Kafka and to get a simple test-case working with SSL authentication, but no authorization (there is no need to start Zookeeper as we already have Apache Atlas running, which starts a Zookeeper instance). Next follow the third tutorial to install the Apache Ranger admin service, as well as the Ranger plugin for Kafka. Create ("resource-based") authorization policies for the Kafka "test" topic in Apache Ranger. There is just one thing we need to change, call the Ranger service "cl1_kafka" instead of "KafkaTest" (this change needs to happen in Ranger, and in the "install.properties" when installing the Ranger plugin to Kafka).

Now verify that the producer has permission to publish to the topic, and the consumer has permission to consume from the topic. Once this is working, then remove the resource-based policy for the consumer, and verify that the consumer no longer has permission to consume from the topic.

3) Use the Apache Ranger TagSync service to import tags from Atlas into Ranger

To create tag based policies in Apache Ranger, we have to import the entity + tag we have created in Apache Atlas into Ranger via the Ranger TagSync service. After building Apache Ranger then extract the file called "target/ranger-<version>-tagsync.tar.gz". There are three alternatives available where the Ranger TagSync service can obtain tag information. From Apache Atlas via a Kafka topic, from Apache Atlas via the REST API and from a file. We will use the REST API of Atlas here. Edit 'install.properties' as follows:
  • Set TAG_SOURCE_ATLAS_ENABLED to "false"
  • Set TAG_SOURCE_ATLASREST_ENABLED to  "true"
  • Set TAG_SOURCE_ATLASREST_DOWNLOAD_INTERVAL_IN_MILLIS to "60000" (just for testing purposes)
  • Specify "admin" for both TAG_SOURCE_ATLASREST_USERNAME and TAG_SOURCE_ATLASREST_PASSWORD
Save 'install.properties' and install the tagsync service via "sudo ./setup.sh". It can now be started via "sudo ranger-tagsync-services.sh start".

4) Create Tag-based authorization policies in Apache Ranger

Now we can create tag-based authorization policies in Apache Ranger. Earlier we used the name "cl1_kafka" for the service name instead of "KafkaTest" as in the previous tutorial. The reason for this is that the service name must match the qualified name attribute of the Kafka entity that we are syncing into Ranger.

In Ranger, click on "Access Manager" and "Tag Based Policies". Create a new "TAG" service called "KafkaTagService". When this is done go into the new service and click on "Add New Policy". Hit (upper-case) "K" in the "TAG" field and "KafkaTag" should pop up automatically (hence the import of tags from Atlas was successful). Add an "allow condition" for the client user with permissions to "consume" and "describe" for "kafka" as shown in the following picture:


Finally, edit the "cl1_kafka" service we created and for "Select Tag Service" select "KafkaTagService" and save. Finally, wait some time for the Ranger plugin to download the new policies and tags and try the consumer again. This time it should work! So we have shown how Ranger can create authorization policies based on tags as well as resources.
Categories: Colm O hEigeartaigh

Securing Apache Atlas using Apache Ranger

Colm O hEigeartaigh - Mon, 02/06/2017 - 13:41
Apache Atlas, currently in the Apache Incubator, is a data governance and metadata framework for Apache Hadoop. It allows you to import data from a backend such as Apache Hive or Apache Falcon, and to classify and tag the data according to a set of business rules. In this tutorial we will show how to to use Apache Ranger to create authorization policies to secure access to Apache Atlas.

1) Set up Apache Atlas

First let's look at setting up Apache Atlas. Download the latest released version (0.7.1-incubating) and extract it. Build the distribution that contains an embedded HBase and Solr instance via:
  • mvn clean package -Pdist,embedded-hbase-solr -DskipTests
The distribution will then be available in 'distro/target/apache-atlas-0.7.1-incubating-bin'. To launch Atlas, we need to set some variables to tell it to use the local HBase and Solr instances:
  • export MANAGE_LOCAL_HBASE=true
  • export MANAGE_LOCAL_SOLR=true
Before starting Atlas, for testing purposes let's add a new user called 'alice' in the group 'DATA_SCIENTIST' with password 'password'. Edit 'conf/users-credentials.properties' and add:
  • alice=DATA_SCIENTIST::5e884898da28047151d0e56f8dc6292773603d0d6aabbdd62a11ef721d1542d8
Now let's start Apache Atlas with 'bin/atlas_start.py'. The Apache Atlas web service can be explored via 'http://localhost:21000/'. To populate some sample data in Apache Atlas, run the command 'bin/quick_start.py' (using credentials admin/admin). To see all traits/tags that have been created, use Curl as follows:
  • curl -u alice:password http://localhost:21000/api/atlas/types?type=TRAIT
2) Install the Apache Ranger Atlas plugin

To use Apache Ranger to secure Apache Atlas, the next step we need to do is to configure and install the Apache Ranger Atlas plugin. Follow the steps in an earlier tutorial to build Apache Ranger and to setup and start the Apache Ranger Admin service. I recommend to use the latest SNAPSHOT of Ranger (0.7.0-SNAPSHOT at this time) as there are some bugs fixed in relation to Atlas support since the 0.6.x release. Once this is done, go back to the Apache Ranger distribution that you have built and extract the atlas plugin:
  • tar zxvf target/ranger-0.7.0-SNAPSHOT-atlas-plugin.tar.gz
 Edit 'install.properties' with the following changes:
  • POLICY_MGR_URL=http://localhost:6080
  • Specify location for SQL_CONNECTOR_JAR 
  • Specify REPOSITORY_NAME (AtlasTest)
  • COMPONENT_INSTALL_DIR_NAME pointing to your Atlas install
Now install the plugin via 'sudo ./enable-atlas-plugin.sh'. If you see an error about "libext" then create a new empty directory called "libext" in the Atlas distribution and try again. Note that the ranger plugin will try to store policies by default in "/etc/ranger/AtlasTest/policycache". As we installed the plugin as "root" make sure that this directory is accessible to the user that is running Apache Atlas. Now restart Apache Atlas to enable the Ranger plugin.

3) Creating authorization policies for Atlas in the Ranger Admin Service

Now that we have set up Apache Atlas to use Apache Ranger for authorization, what remains is to start the Apache Ranger Admin Service and to create some authorization policies. Start Apache Ranger ('sudo ranger-admin start'). Log in to 'http://localhost:6080/' (credentials admin/admin). Click on the "+" button for Atlas, and specify the following fields:
  • Service Name: AtlasTest
  • Username: admin
  • Password: admin
  • atlas.rest.address: http://localhost:21000
Click on "Test Connection" to make sure that we can communicate successfully with Apache Atlas and then "Add". Click on the new link for "AtlasTest". Let's see if our new user "alice" is authorized to read the tags in Atlas. Execute the Curl command defined above (allowing 30 seconds for the Ranger plugin to pull the policies from the Ranger Admin Service). You should see a 403 Forbidden message from Atlas.

Now let's update the authorization policies to allow "alice" access to reading the tags. Back in Apache Ranger, click on "Settings" and then "Users/Groups" and "Groups". Click on "Add new group" and enter "DATA_SCIENTIST" for the name. Now go back into "AtlasTest", and edit the policy called "all - type". Create a new "Allow Condition" for the group "DATA_SCIENTIST" with permission "read" and click "Save". After waiting some time for the policies to sync, try again with the "Curl" command and it should work.


Categories: Colm O hEigeartaigh

Authenticating users in the Apache Ranger Admin Service via PAM

Colm O hEigeartaigh - Thu, 02/02/2017 - 13:08
Over the past few months, I've written various tutorials about different ways you can authenticate to the Apache Ranger Admin Service. In summary, here are the options that have been covered so far:
The remaining option is to authenticate users directly to the local UNIX machine. There is a legacy way of doing this that supports authentication using shadow files. However, a much better approach is to support user authentication using Pluggable Authentication Modules (PAM). This means we can delegate user authentication to various PAM modules, and so we have a wide range of user authentication options. In this post we will show how to configure the Ranger Admin Service to authenticate users on a local linux machine using PAM. There is also an excellent in-depth tutorial that covers PAM and Ranger available here.

1) Configuring the Apache Ranger Admin Service to use PAM for authentication

Follow the steps in a previous tutorial to build Apache Ranger and to setup and install the Apache Ranger Admin service. Edit 'conf/ranger-admin-site.xml' and change the following configuration value:
  • ranger.authentication.method: PAM
2) Add a PAM configuration file for Apache Ranger

The next step is to add a PAM configuration file for Apache Ranger. Create a file called '/etc/pam.d/ranger-admin' with the content:
  • auth    required    pam_unix.so
  • account    required    pam_unix.so
Essentially this means that we are delegating authentication to the local unix machine. Now start the Apache Ranger Admin service. You should be able to log on to http://localhost:6080/login.jsp using a local user credential.
Categories: Colm O hEigeartaigh

Syncing Users and Groups from UNIX into Apace Ranger

Colm O hEigeartaigh - Tue, 01/17/2017 - 13:40
The previous blog post showed how to authenticate users logging in to the Apache Ranger admin service via LDAP. An older blog post covered how to sync users and groups from LDAP into Apache Ranger so that they can be used both for authentication and to construct authorization policies. Another option is to sync users and groups from the local UNIX machine into Apache Ranger, something we will cover in this post.

1) Build the Apache Ranger usersync module

Follow the steps in the following tutorial to build Apache Ranger and to setup and start the Apache Ranger Admin service. Once this is done, go back to the Apache Ranger distribution that you have built and copy the usersync module:
  • tar zxvf target/ranger-0.6.0-usersync.tar.gz
  • mv ranger-0.6.0-usersync ${usersync.home}
2) Configure and build the Apache Ranger usersync service 

You will need to install the Apache Ranger Usersync service using "sudo". If the root user does not have a JAVA_HOME property defined, then edit ${usersync.home}/setup.sh + add in, e.g.:
  • export JAVA_HOME=/opt/jdk1.8.0_112
Next edit ${usersync.home}/install.properties and make the following changes:
  • POLICY_MGR_URL = http://localhost:6080
  • SYNC_SOURCE = unix
  • SYNC_INTERVAL = 1 (just for testing purposes....)
Now you can run the setup script via "sudo ./setup.sh". 

3) Start the Usersync service

The Apache Ranger Usersync service can be started via "sudo ./ranger-usersync-services.sh start". After 1 minute (see SYNC_INTERVAL above), it should successfully copy the users/groups from the local UNIX machine into the Apache Ranger Admin. Open a browser and go to "http://localhost:6080", and click on "Settings" and then "Users/Groups". You should see the users and groups synced successfully.
 

Categories: Colm O hEigeartaigh

Authenticating users in the Apache Ranger Admin Service via LDAP

Colm O hEigeartaigh - Wed, 12/21/2016 - 18:17
In the summer, I wrote a couple of blog posts covering how to configure and install the Apache Ranger Admin Service and then how to use the Apache Ranger usersync Service to import users and groups from LDAP into the Apache Ranger Admin Service. The Usersync service periodically imports all users and groups that matches the configured search base into the Apache Ranger Admin service. However, this bulk import of users and groups might be unnecessary for your particular requirements. In this post we will show instead how to authenticate users logging in to the Apache Ranger Admin Service UI using LDAP.

1) The OpenDS backend

As with the tutorial on the Apache Ranger usersync service, we will use OpenDS as the LDAP server. It contains a domain called "dc=example,dc=com", and 5 users (alice/bob/dave/oscar/victor) and 2 groups (employee/manager). Victor, Oscar and Bob are employees, Alice and Dave are managers. Here is a screenshot using Apache Directory Studio:


2) Configuring the Apache Ranger Admin Service to use LDAP for authentication

Follow the steps in the following tutorial to build Apache Ranger and to setup and install the Apache Ranger Admin service. Edit 'conf/ranger-admin-site.xml' and change/specify the following configuration values:
  • ranger.authentication.method: LDAP
  • ranger.ldap.url: ldap://localhost:2389
  • ranger.ldap.user.dnpattern: cn={0},ou=users,dc=example,dc=com
  • ranger.ldap.group.searchbase: ou=groups,dc=example,dc=com
  • ranger.ldap.group.searchfilter: (member=cn={1},ou=users,dc=example,dc=com)
  • ranger.ldap.group.roleattribute: cn
  • ranger.ldap.base.dn: dc=example,dc=com
  • ranger.ldap.bind.dn: cn=Directory Manager,dc=example,dc=com
  • ranger.ldap.bind.password: test
Note that the "group" configuration attributes must be specified, even though the group information is not actually used. I've submitted a patch for this which should be fixed for the next Ranger release. Now simply save the changes to 'conf/ranger-admin-site.xml' and start the Apache Ranger Admin service. You should be able to log on to http://localhost:6080/login.jsp using the LDAP credentials store in OpenDS.
Categories: Colm O hEigeartaigh

Home Realm Discovery in the Apache CXF Fediz IdP

Colm O hEigeartaigh - Fri, 11/11/2016 - 17:42
When a client application (secured via either WS-Federation or SAML SSO) redirects a user to the Apache CXF Fediz IdP, the IdP must figure out what the home realm of the user is. If the home realm of the user corresponds to the realm of the IdP, then the IdP can authenticate the user. However, if the home realm does not match that of the IdP, then the IdP has the option to forward the authentication request to a third party IdP for authentication, if it is configured to do this. In this post, we will look at the different options available in the IdP to figure out what the home realm of the user is.

1) The 'whr' query parameter

When using the WS-Federation protocol, the application can specify the home realm of the user by adding the 'whr' query parameter to the URI that the browser is redirected to. Alternatively, the 'whr' query parameter could be added by a reverse proxy sitting in front of the IdP. Here is an example of such a URI including a 'whr' query parameter:
  • https://localhost:45753/fediz-idp-realmb/federation?wa=wsignin1.0&wtrealm=urn%3Aorg%3Aapache%3Acxf%3Afediz%3Aidp%3Arealm-A&wreply=https%3A%2F%2Flocalhost%3A43618%2Ffediz-idp%2Ffederation&whr=urn:org:apache:cxf:fediz:idp:realm-B&wfresh=10&wctx=c07a5b9a-e270-4855-9201-fc1801851cc9
2) The 'hrds' configuration option in the IdP

If no 'whr' query parameter is available (this will always be the case for SAML SSO), then the IdP attempts to find out the home realm of the user by querying the "hrds" property of the IdP. This is a Spring Expression Language expression that is evaluated on the Spring WebFlow RequestContext.

For an example of how this can be used, let's look at the tests in Fediz for the SAML SSO IdP when redirecting to a trusted third party IdP. As there is no 'whr' query parameter for SAML SSO, instead we will define a class with a static method that maps application realms to home realms. The application realm is available in the IdP, as the SAML SSO AuthnRequest is already parsed at this point (it corresponds to the "Issuer" of the AuthnRequest). So we can specify the hrds configuration options in the IdP as follows:
  • <property name="hrds" value="T(org.apache.cxf.fediz.integrationtests.RealmMapper).realms()                                   .get(getFlowScope().get('saml_authn_request').issuer)" />
3) Via a form

If no 'whr' query parameter is available, and no 'hrds' configuration option is specified, then the IdP will display a form where the user can select the home realm. The IdP only does this if the configuration option "provideIdpList" is set to true. If it is set to false, then the current IdP is assumed to be the home realm IdP, unless the configuration option "useCurrentIdp" is also set to "false", in which case an error is displayed. The user can select the home realm in the form corresponding to the known trusted IdP realms of this IdP:


Categories: Colm O hEigeartaigh

Support for IdP-initiated SAML SSO in Apache CXF

Colm O hEigeartaigh - Fri, 11/04/2016 - 18:26
Previous blog posts have covered how to secure your JAX-RS web applications in Apache CXF using SAML SSO. Since the 3.1.8 release, Apache CXF also supports IdP-initiated SAML SSO. The typical use-case for SAML SSO involves the browser invoking on a JAX-RS application, and then being redirected to an IdP for authentication, which subsequently redirects the browser back to the application. However, sometimes a user will log on first to the IdP and then want to invoke on a web application. In this post we will show how to configure SAML SSO for a CXF-based web application to support the IdP-initiated flow, by demonstrating an interop test-case with Okta.

1) Configuring a SAML application in Okta

The first step is to create an account at Okta and configure a SAML application. This process is mapped out at the following link. Follow the steps listed on this page with the following additional changes:
  • Specify the following for the Single Sign On URL and audience URI: http://localhost:8080/fedizdoubleit/racs/sso
  • Specify the following for the default RelayState: http://localhost:8080/fedizdoubleit/app1/services/25
  • Add an additional attribute with name "http://schemas.xmlsoap.org/ws/2005/05/identity/claims/role" and value "Manager".
The RequestAssertionConsumerService will process the SAML Response from Okta. However, it doesn't know where to subsequently send the browser. Therefore, we are configuring the RelayState parameter to encode the URL of the end application. In addition, our test application requires that the user has a specific role to invoke upon it, hence we add a "Manager" attribute with the URI corresponding to a role.

When the application is configured, you will see an option to "View Setup Instructions". Open this link in a new tab and set it aside for the moment - it contains information required when setting up the web application. Now click on the "People" tab and assign the application to the username that you have created at Okta.

2) Setting up the SAML SSO-enabled web application

We will use a trivial "double it" web application which I wrote previously to demonstrate the SAML SSO capabilities of Apache CXF Fediz. The web application is available here. Build the web application and deploy it in Apache Tomcat. You will need to edit 'webapps/fedizdoubleit/WEB-INF/cxf-service.xml'.

a) SamlRedirectBindingFilter configuration changes

First let's look at the changes which are required to the 'SamlRedirectBindingFilter':
  • Remove "idpServiceAddress" and "assertionConsumerServiceAddress". These aren't required as we are only supporting the IdP-initiated flow.
  • Also remove the "signRequest", "signaturePropertiesFile", "callbackHandler", "signatureUsername" and "issuerId" properties.
  • Add <property name="addWebAppContext" value="false"/>
  • Add <property name="supportUnsolicited" value="true"/>

b) RequestAssertionConsumerService (RACS) configuration changes

Now add the following properties to the "RequestAssertionConsumerService":
  • <property name="supportUnsolicited" value="true"/>
  • <property name="idpServiceAddress" value="..."/>
  • <property name="issuerId" value="http://localhost:8080/fedizdoubleit/racs/sso"/>
  • <property name="parseApplicationURLFromRelayState" value="true"/>
Paste in the value for "idpServiceAddress" from the "Identity Provider Single Sign-On URL" given in the "View Setup Instructions" page referenced above.
c) Adding Okta cert into the RACS truststore

As things stand, the SAML Response from Okta to the RequestAssertionConsumerService will fail, as the RACS will not trust the certificate Okta uses to sign the SAML Response. Therefore we need to insert the Okta cert into the truststore of the RACS. Copy the "X.509 Certificate" value from the "View Setup Instructions" page referenced earlier. Create a file called 'webapps/fedizdoubleit/WEB-INF/classes/okta.cert' and paste the certificate contents into this file. Import it into the truststore via:
  • keytool -keystore stsrealm_a.jks -storepass storepass -importcert -file okta.cert
At this point we should be all done. Click on the box for the application you have created in Okta. You should be redirected to the CXF RACS, which validates the SAML Response, and in turn redirects to the application.


Categories: Colm O hEigeartaigh

Client Credentials grant support in the Apache CXF Fediz OIDC service

Colm O hEigeartaigh - Wed, 11/02/2016 - 18:59
Apache CXF Fediz ships with a powerful and flexible OpenId Connect (OIDC) service that can be used to implement SSO for your organisation. All of the core OIDC flows are supported - Authorization Code flow, Implicit and Hybrid flows. As OIDC is just an identity layer over OAuth 2.0, it's possible to use Fediz as a purely OAuth 2.0 service as well, and all of the authorization grants defined in the spec are also fully supported. In this post we will look at support for one of these authorization grants in Fediz 1.3.1 - the client credentials grant.

1) The OAuth 2.0 client credentials grant

The client credentials grant is used for when the client is requesting access for a resource that is owned or controlled by that client. There is no enduser in this scenario, unlike say the authorization code flow or implicit flow. The client simply calls the token endpoint of the authorization service using "client_credentials" for the "grant_type" parameter. In addition, the client must authenticate (e.g. by supplying client_id and client_secret parameters). The authorization service authenticates the client and then returns an access token.

2) Supporting the client credentials grant in Fediz OIDC

It's easy to support the client credentials grant in the Fediz OIDC service.

a) Add the ClientCredentialsGrantHandler

Firstly, the ClientCredentialsGrantHandler must be added to the list of grant handlers supported by the token service as follows:

b) Add a way of authenticating the client

The next step is to add a way of authenticating the client credentials. Fediz uses JAAS to make it easy for the deployer to plug in different JAAS LoginModules if required. To configure JAAS, you must specify the name of the JAAS LoginModule in the configuration of the OAuthDataProviderImpl:

c) Example JAAS configuration

For the normal OIDC flows, the Fediz OIDC uses a WS-Federation filter to redirect the browser to the Fediz IdP, where the end user is then ultimately authenticated by the STS that bundles with Fediz. Therefore it seems like a natural fit to re-use the STS to authenticate the client in the Fediz OIDC. Follow steps (a) and (b) above. Start the Fediz STS, but before starting the OIDC service, specify the "java.security.auth.login.config" system property to point to the following JAAS configuration file:

You must substitute the correct port for "${idp.https.port}". The STSLoginModule takes the given username and password supplied by the client and uses them to authenticate to the STS.
Categories: Colm O hEigeartaigh

Switching authentication mechanisms in the Apache CXF Fediz STS

Colm O hEigeartaigh - Wed, 10/26/2016 - 17:56
Apache CXF Fediz ships with an Identity Provider (IdP) that can authenticate users via either the WS-Federation or SAML SSO protocols. The IdP delegates user authentication to a Security Token Service (STS) web application using the WS-Trust protocol. The STS implementation in Fediz ships with some sample user data for use in the tests. For a real-world scenario, deployers will have to swap the sample data out for an identity backend (such as Active Directory or LDAP). This post will explain how this can be done, with a particular focus on some recent changes to the STS web application in Fediz to make the process easier.

1) The default STS that ships with Fediz

First let's explain a bit about how the STS is configured by default in Fediz to cater for the testcases.

a) Endpoints and user authentication

The STS must define two distinct set of endpoints to work with the IdP. Firstly, the STS must be able to authenticate the user credentials that are presented to the IdP. Typically this is a Username + Password combination. However, X.509 client certificates and Kerberos tokens are also supported. Note that by default, the STS authenticates usernames and passwords via a simple file local to the STS.

After successful user authentication, a SAML token is returned to the IdP. The IdP then gets another SAML token "on behalf of" the authenticated user for a given realm, authenticating using its own credentials. So we need a second endpoint in the STS to issue this token. By default, the STS requires that the IdP authenticate using TLS client authentication. The security policies are defined in the WSDLs available here.

b) Realms

The Fediz IdP and STS support the concept of authenticating users in different realms. By default, the IdP is configured to authenticate users in "Realm A". This corresponds to a specific endpoint address in the STS. The STS also defines user authentication endpoints in "Realm B" for use in test scenarios involving identity federation between two IdPs.

In addition, the STS defines some configuration to map user identities between realms. In other words, how a principal in one realm should map to another realm, and how the claims in one realm map to those in another realm.

2) Changing the STS in Fediz 1.3.2 to use LDAP

From the forthcoming 1.3.2 release onwards, the Fediz STS web application is a bit easier to customize for your specific deployment needs. Let's see how easy it is to switch the STS to use LDAP.

a) Deploy the vanilla IdP and STS to Apache Tomcat

To start with, we will deploy the STS and IdP containing the sample data to Apache Tomcat.
  • Create a new directory: ${catalina.home}/lib/fediz
  • Edit ${catalina.home}/conf/catalina.properties and append ',${catalina.home}/lib/fediz/*.jar' to the 'common.loader' property.
  • Copy ${fediz.home}/plugins/tomcat/lib/* to ${catalina.home}/lib/fediz
  • Copy ${fediz.home}/idp/war/* to ${catalina.home}/webapps
  • Download and copy the hsqldb jar (e.g. hsqldb-2.3.4.jar) to ${catalina.home}/lib 
  • Copy idp-ssl-key.jks and idp-ssl-trust.jks from ${fediz.home}/examples/sampleKeys to ${catalina.home}
  • Edit the TLS Connector in ${catalina.home}/conf/server.xml', e.g.: <Connector port="8443" protocol="org.apache.coyote.http11.Http11Protocol" maxThreads="150" SSLEnabled="true" scheme="https" secure="true" clientAuth="want" sslProtocol="TLS" keystoreFile="idp-ssl-key.jks" keystorePass="tompass" keyPass="tompass" truststoreFile="idp-ssl-trust.jks" truststorePass="ispass" />
Now start Tomcat and then enter the following in a web browser (authenticating with "alice/ecila" in "realm A" - you should be directed to the URL for the default service application (404, as we have not configured it):

https://localhost:8443/fediz-idp/federation?wa=wsignin1.0&wreply=https%3A%2F%2Flocalhost%3A8443%2Ffedizhelloworld%2Fsecure%2Ffedservlet&wtrealm=urn%3Aorg%3Aapache%3Acxf%3Afediz%3Afedizhelloworld

b) Change the STS authentication mechanism to Active Directory

To simulate an Active Directory instance for demonstration purposes, we will modify some LDAP system tests in the Fediz source that use Apache Directory. Check out the Fediz source and build it via "mvn install -DskipTests". Now go into "systests/ldap" and edit the LDAPTest. "@Ignore" the existing test + uncomment the test which just "sleeps". Also change the "@CreateTransport" annotation to start the LDAP port on "12345" instead of a random port.

Next we'll configure the Fediz STS to use this LDAP instance for authentication. Edit 'webapps/fediz-idp-sts/WEB-INF/cxf-transport.xml'. Change "endpoints/file.xml" to "endpoints/ldap.xml". Next edit 'webapps/fediz-idp-sts/WEB-INF/endpoints/ldap.xml" and just change the port from "389" to "12345".

Now we need to configure a JAAS configuration file, which the STS uses to validate the received Username + Password to LDAP. Copy this file to the "conf" directory of Tomcat, substituting "12345" for "portno". Now restart Tomcat, this time specifying the location of the JAAS configuration file, e.g.:
  • export JAVA_OPTS="-Xmx2048M -Djava.security.auth.login.config=/opt/fediz-apache-tomcat-8.0.37/conf/ldap.jaas"
This is all the changes that are required to swap over to use an LDAP instance for authentication.
    Categories: Colm O hEigeartaigh

    Securing an Apache Kafka broker - part IV

    Colm O hEigeartaigh - Wed, 09/28/2016 - 13:24
    This is the fourth in a series of articles on securing an Apache Kafka broker. The first post looked at how to secure messages and authenticate clients using SSL. The second post built on the first post by showing how to perform authorization using some custom logic. The third post showed how Apache Ranger could be used instead to create and enforce authorization policies for Apache Kafka. In this post we will look at an alternative authorization solution called Apache Sentry.

    1) Build the Apache Sentry distribution

    First we will build and install the Apache Sentry distribution. Download Apache Sentry (1.7.0 was used for the purposes of this tutorial). Verify that the signature is valid and that the message digests match. Now extract and build the source and copy the distribution to a location where you wish to install it:
    • tar zxvf apache-sentry-1.7.0-src.tar.gz
    • cd apache-sentry-1.7.0-src
    • mvn clean install -DskipTests
    • cp -r sentry-dist/target/apache-sentry-1.7.0-bin ${sentry.home}
    Apache Sentry has an authorization plugin for Apache Kafka, amongst other big data projects. In addition it comes with an RPC service which stores authorization privileges in a database. For the purposes of this tutorial we will just configure the authorization privileges in a configuration file locally to the broker. Therefore we don't need to do any further configuration to the distribution at this point.

    2) Configure authorization in the broker

    Configure Apache Kafka as per the first tutorial. To enable authorization using Apache Sentry we also need to follow these steps. First edit 'config/server.properties' and add:
    • authorizer.class.name=org.apache.sentry.kafka.authorizer.SentryKafkaAuthorizer
    • sentry.kafka.site.url=file:./config/sentry-site.xml
    Next copy the jars from the "lib" directory of the Sentry distribution to the Kafka "libs" directory. Then create a new file in the config directory called "sentry-site.xml" with the following content:

    This is the configuration file for the Sentry plugin for Kafka. It essentially says that the authorization privileges are stored in a local file, and that the groups for authenticated users should be retrieved from this file. Finally, we need to specify the authorization privileges. Create a new file in the config directory called "sentry.ini" with the following content:

    This configuration file contains three separate sections. The "[users]" section maps the authenticated principals to local groups. The "[groups]" section maps the groups to roles, and the "[roles]" section lists the actual privileges. Now we can start the broker as in the first tutorial:
    • bin/kafka-server-start.sh config/server.properties 
    3) Test authorization

    Now lets test the authorization logic. Start the producer:
    • bin/kafka-console-producer.sh --broker-list localhost:9092 --topic test --producer.config config/producer.properties
    Send a few messages to check that the producer is authorized correctly. Now start the consumer:
    • bin/kafka-console-consumer.sh --bootstrap-server localhost:9092 --topic test --from-beginning --consumer.config config/consumer.properties --new-consumer
    If everything is configured correctly then it should work as in the first tutorial. 
    Categories: Colm O hEigeartaigh

    Securing an Apache Kafka broker - part III

    Colm O hEigeartaigh - Mon, 09/26/2016 - 18:13
    This is the third in a series of blog posts about securing Apache Kafka. The first post looked at how to secure messages and authenticate clients using SSL. The second post built on the first post by showing how to perform authorization using some custom logic. However, this approach is not recommended for non-trivial deployments. In this post we will show at how we can create flexible authorization policies for Apache Kafka using the Apache Ranger admin UI. Then we will show how to enforce these policies at the broker.

    1) Install the Apache Ranger Kafka plugin

    The first step is to download Apache Ranger (0.6.1-incubating was used in this post). Verify that the signature is valid and that the message digests match. Now extract and build the source, and copy the resulting plugin to a location where you will configure and install it:
    • tar zxvf apache-ranger-incubating-0.6.1.tar.gz
    • cd apache-ranger-incubating-0.6.1
    • mvn clean package assembly:assembly -DskipTests
    • tar zxvf target/ranger-0.6.1-kafka-plugin.tar.gz
    • mv ranger-0.6.1-kafka-plugin ${ranger.kafka.home}
    Now go to ${ranger.kafka.home} and edit "install.properties". You need to specify the following properties:
    • COMPONENT_INSTALL_DIR_NAME: The location of your Kafka installation
    • POLICY_MGR_URL: Set this to "http://localhost:6080"
    • REPOSITORY_NAME: Set this to "KafkaTest".
    Save "install.properties" and install the plugin as root via "sudo ./enable-kafka-plugin.sh". The Apache Ranger Kafka plugin should now be successfully installed (although not yet configured properly) in the broker.

    2) Configure authorization in the broker

    Configure Apache Kafka as per the first tutorial. There are a number of steps we need to follow to configure the Ranger Kafka plugin before it is operational:
    • Edit 'config/server.properties' and add the following: authorizer.class.name=org.apache.ranger.authorization.kafka.authorizer.RangerKafkaAuthorizer
    • Add the Kafka "config" directory to the classpath, so that we can pick up the Ranger configuration files: export CLASSPATH=$KAFKA_HOME/config
    • Copy the Apache Commons Logging jar into $KAFKA_HOME/libs. 
    • The ranger plugin will try to store policies by default in "/etc/ranger/KafkaTest/policycache". As we installed the plugin as "root" make sure that this directory is accessible to the user that is running the broker.
    Now we can start the broker as in the first tutorial:
    • bin/kafka-server-start.sh config/server.properties
    3) Configure authorization policies in the Apache Ranger Admin UI 

    At this point we should have configured the broker so that the Apache Ranger plugin is used to communicate with the Apache Ranger admin service to download authorization policies. So we need to install and configure the Apache Ranger admin service. Please refer to this blog post for how to do this. Assuming the admin service is already installed, start it via "sudo ranger-admin start". Open a browser and log on to "localhost:6080" with the credentials "admin/admin".

    First lets add some new users that match the SSL principals we have created in the first tutorial. Click on "Settings" and "Users/Groups". Add new users for the principals:
    • CN=Client,O=Apache,L=Dublin,ST=Leinster,C=IE
    • CN=Service,O=Apache,L=Dublin,ST=Leinster,C=IE
    • CN=Broker,O=Apache,L=Dublin,ST=Leinster,C=IE
    Now go back to the Service Manager screen and click on the "+" button next to "KAFKA". Create a new service called "KafkaTest". Click "Test Connection" to make sure it can communicate with the Apache Kafka broker. Then click "add" to save the new service. Click on the new service. There should be an "admin" policy already created. Edit the policy and give the "broker" principal above the rights to perform any operation and save the policy. Now create a new policy called "TestPolicy" for the topic "test". Give the service principal the rights to "Consume, Describe and Publish". Give the client principal the rights to "Consum and Describe" only.


    4) Test authorization

    Now lets test the authorization logic. Bear in mind that by default the Kafka plugin reloads policies from the admin service every 30 seconds, so you may need to wait that long or to restart the broker to download the newly created policies. Start the producer:
    • bin/kafka-console-producer.sh --broker-list localhost:9092 --topic test --producer.config config/producer.properties
    Send a few messages to check that the producer is authorized correctly. Now start the consumer:
    • bin/kafka-console-consumer.sh --bootstrap-server localhost:9092 --topic test --from-beginning --consumer.config config/consumer.properties --new-consumer
    If everything is configured correctly then it should work as in the first tutorial.
    Categories: Colm O hEigeartaigh

    Integrating Apache Camel with Apache Syncope - part III

    Colm O hEigeartaigh - Fri, 09/23/2016 - 18:00
    This is the third in a series of blog posts about integrating Apache Camel with Apache Syncope. The first post introduced the new Apache Camel provisioning manager that is available in Apache Syncope 2.0.0, and gave an example of how we can modify the default behaviour to send an email to an administrator when a user was created. The second post showed how an administrator can keep track of user password changes for auditing purposes. In this post we will show how to integrate Syncope with Apache ActiveMQ using Camel.

    1) The use-case

    The use-case is that Apache Syncope is used for Identity Management in a large organisation. When users are created we would like to be able to gather certain information about the new users and process it dynamically in some way. In particular, we are interested in the age of the new users and the country in which they are based. Perhaps at the reception desk of the company HQ we display a map with the number of employees in each country highlighted. To decouple whatever applications are processing the data from Syncope itself, we will use a messaging solution, namely Apache ActiveMQ. When new users are created, we will modify the default Camel route to send a message to two topics corresponding to the age and location of the user.

    2) Download and configure Apache ActiveMQ

    The first step is to download Apache ActiveMQ (currently 5.14.0). Unzip it and start it via:
    • bin/activemq start 
    Now go to the web interface of ActiveMQ - 'http://localhost:8161/admin/', logging in with credentials 'admin/admin'. Click on the "Queues" tab and create two new queues called 'age' and 'country'.

    3) Download and configure Apache Syncope

    Download and extract the standalone version of Apache Syncope 2.0.0. Before we start it we will copy the jars we need to get Camel working with ActiveMQ in Syncope. In the "webapps/syncope/WEB-INF/lib" directory of the Apache Tomcat instance bundled with Syncope, copy the following jars:
    • From $ACTIVEMQ_HOME/lib: activemq-client-5.14.0.jar + activemq-spring-5.14.0.jar + hawtbuf-1.11.jar + geronimo-j2ee-management_1.1_spec-1.0.1.jar
    • From $ACTIVEMQ_HOME/lib/camel: activemq-camel-5.14.0.jar + camel-jms-2.16.3.jar
    • From $ACTIVEMQ_HOME/lib/optional: activemq-pool-5.14.0.jar + activemq-jms-pool-5.14.0.jar + spring-jms-4.1.9.RELEASE.jar
    Next we need to create a Camel spring configuration file containing a bean with the address of the broker. Add the following file to the Tomcat lib directory (called "camelRoutesContext.xml"):

    Now we can start the embedded Apache Tomcat instance. Open a browser and navigate to 'http://localhost:9080/syncope-console' logging in with 'admin/password'. The first thing we need to do is to configure user attributes for "age" and "country". Go to "Configuration/Types" in the left-hand menu, and click on the "Schemas" tab. Create two plain (mandatory) schema types: "age" of type "String" and "country" of type "Long". Now click on the "AnyTypeClasses" tab and create a new AnyTypeClass selecting the two plain schema types we just created. Finally, click on the "AnyType" tab and edit the "USER". Add the new AnyTypeClass you created and hit "save".

    Now we will modify the Camel route invoked when a user is created. Click on "Extensions/Camel Routes" in the left-hand configuration menu. Edit the "createUser" route and add the following above the "bean method" part:
    • <setBody><simple>${body.plainAttrMap[age].values[0]}</simple></setBody>
    • <to uri="activemq:age"/>
    • <setBody><simple>${exchangeProperty.actual.plainAttrMap[country].values[0]}</simple></setBody>
    • <to uri="activemq:country"/>
    This should be fairly straightforward to follow. We are setting the message body to be the age of the newly created User, and dispatching that message to the "age" queue. We then follow the same process for the "country". We also need to change "body" in the "bean method" line to "exchangeProperty.actual", this is because we have redefined what the body is for each of the Camel routes above.


    Now let's create some new users. Click on the "Realms" menu and select the "USER" tab. Create new users "alice" in country "usa" of age "25" and "bob" in country "canada" of age "27". Now let's look at the ActiveMQ console again. We should see two new messages in each of the queues as follows, ready to be consumed:



    Categories: Colm O hEigeartaigh

    Pages

    Subscribe to Talend Community Coders aggregator - Colm O hEigeartaigh