Apache Kafka is a distributed, partitioned, replicated commit log service. It provides the functionality of a messaging system, but with a unique design that enables Kafka to achieve very high throughput and very low latencies.
First let's review some basic messaging terminology:
So, at a high level, producers send messages over the network to the Kafka cluster which in turn serves them up to consumers like this:
Starting from 0.9.0.0 release, Apache Kafka has various security features built in, like, encryption and authentication using SSL, authentication using SASL, Apache Zookeeper authentication, quotas, and authorization. There are various ways one would want to have authorization done and so Kafka allows different authorization implementations to plug into Kafka. By default Apache Kafka comes with a Zookeeper based authorization implementation, which uses Zookeeper to store ACLs. Note that Kafka has a hard dependency on Zookeeper for configuration management and leader election, and is not an additional requirement to be able to use out of box Zookeeper based authorization. Though it is nice to not have any dependency on an external system for out-of-the-box authorization implementation in Apache Kafka, it has quite a few shortcomings.
Apache Sentry is a system for enforcing fine grained role based authorization. Role Based Authorization Control, RBAC, is a powerful mechanism to manage authorization for a large set of users and data objects in a typical enterprise. Apache Sentry allows for various systems to integrate with it for utilizing it's generic and powerful authorization. Many systems, like, Hive, Impala, HDFS, Sqoop, etc are already capable of using Apache Sentry for providing authorization. It is also capable of getting user group mapping from external systems, like, LDAP, AD, etc. All the shortcomings of Zookeeper based out-of-the-box Apache Kafka authorization implementation can be taken care of if we choose Apache Sentry to provide authorization in Apache Kafka as well.
Starting from 1.7.0 release, Apache Sentry has Kafka binding that can be used to enable authorization in Apache Kafka with Apache Sentry. Following sections go over how to configure Apache Kafka to use Apache Sentry for authorization and a quick-start guide.
To enable authorization in Apache Kafka and use Apache Sentry for authorization, follow these steps.
This tutorial assumes you have a working Kafka and Sentry installations, Kafka is configured to use Sentry for authorization, and you are starting fresh and have no existing Kafka or ZooKeeper data.
org.apache.sentry.provider.db.generic.tools.SentryShellKafka
is a tool that takes a configuration file with connection details of Sentry server. One can simply wrap this in a shell script or create an alias for easy access. Below is a sample shell script that uses Apache Kafka's kafka-run-class script that takes care of building classpath. Assuming you have copied all required Sentry jars into Kafka's libs dir or have added those JARS to CLASSPATH
env variable, which is used by kafka-run-class script, below script will work for you.
#!/bin/bash # Licensed to the Apache Software Foundation (ASF) under one or more # contributor license agreements. See the NOTICE file distributed with # this work for additional information regarding copyright ownership. # The ASF licenses this file to You under the Apache License, Version 2.0 # (the "License"); you may not use this file except in compliance with # the License. You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. CONFIG_OPTION="--config" KAFKA_CONF_DIR=${KAFKA_CONF_DIR:-/etc/kafka/conf} SENTRY_CONF_DIR=${SENTRY_CONF_DIR:-$KAFKA_CONF_DIR/sentry-conf} SENTRY_SITE_XML=${SENTRY_SITE_XML:-sentry-site.xml} SENTRY_CONF_FILE=$SENTRY_CONF_DIR/$SENTRY_SITE_XML USAGE_STRING="USAGE: kafka-sentry [$CONFIG_OPTION <path_to_sentry_conf_dir>] <sentry-cli-arguments>" if [[ "$CONFIG_OPTION" == $1 ]]; then conf_dir=$2 shift;shift conf_file=$conf_dir/$SENTRY_SITE_XML if [[ ! -f $conf_file ]]; then echo "Configuration file, ${conf_file}, does not exist." echo "${USAGE_STRING}" exit 1 fi else if [[ -f "${SENTRY_CONF_FILE}" ]]; then conf_file=${SENTRY_CONF_FILE} else echo "No configuration directory for Sentry specified and default conf file ${SENTRY_CONF_FILE} does not exist. Please provide a configuration directory that contains sentry-site.xml with information on how to connect with Sentry service." echo "${USAGE_STRING}" exit 1 fi fi # supress the HADOOP_HOME warnings in 1.x.x export HADOOP_HOME_WARN_SUPPRESS=true exec $(dirname $0)/kafka-run-class.sh org.apache.sentry.provider.db.generic.tools.SentryShellKafka -conf $conf_file "$@" |
Rest of this document will refer this script as kafka-sentry
and will use it for performing CRUD operations on roles and privileges. Below is usage info from the tool.
$ kafka-sentry usage: kafka-sentry -arg,--add_role_group Add role to group -conf,--sentry_conf <arg> sentry-site file path -cr,--create_role Create role -dr,--drop_role Drop role -drg,--delete_role_group Delete role from group -g,--groupname <arg> Group name -gpr,--grant_privilege_role Grant privilege to role -h,--help Shell usage -lp,--list_privilege List privilege -lr,--list_role List role -p,--privilege <arg> Privilege string -r,--rolename <arg> Role name -rpr,--revoke_privilege_role Revoke privilege from role |
Here we create a role, test
.
$ kafka-sentry -cr -r test $ kafka-sentry -lr test |
Here we assign the created role, test,
to a group, test-group
. All users in this group, will get any privilege we grant to the role, test
.
$ kafka-sentry -arg -r test -g test-group |
Authorizable resources are resources or entities in a Kafka cluster that require special permissions for a user to be able to perform some action on it. As of now Kafka has three authorizable resources.
kafka-cluster
, as one Kafka cluster can not have more than one Cluster resources.group.id
of a consumer group.*
, as a value, which represents all hosts.Each resource can have multiple actions that users can perform on them. Following operations are supported in Kafka, however not all actions are valid on all resources.
Privileges define what actions are allowed on a resource. A privilege is represented as a string in Sentry. Following are the criterias of a valid privilege.
Host
resource. If no Host
resource is specified in a privilege string, Host=*
is assumed.Host
resource.Valid privilege strings
Host=*->Topic=test->action=ALL
Topic=test->action=ALL
Invalid privilege strings
Host=*->Host=127.0.0.1->Topic=test->action=ALL
, has multiple Host
resourcesCluster=kafka-cluster->Topic=test->action=ALL
, has multiple non Host
resourcesTopic=test->action=ALL->action=read
, has multiple actionsCluster=cluster1->Topic=test->action=ALL
, should only have kafka-cluster
as Cluster
valueaction=ALL->Topic=test
, action must be specified at the end of privilege stringHere we grant some privileges to the role, test, so that users in
testGroup
can create a topic, testTopic
, and produce to it.
Allow users in test-group
to create a new topic from localhost.
$ kafka-sentry -gpr -r test -p "Host=127.0.0.1->Cluster=kafka-cluster->action=create" |
Allow users in test-group
to describe testTopic from localhost, which the user will create and use
.
$ kafka-sentry -gpr -r test -p "Host=127.0.0.1->Topic=testTopic->action=describe" |
Allow users in test-group
to write to testTopic
from localhost, this will allow the users to produce to testTopic
.
$ kafka-sentry -gpr -r test -p "Host=127.0.0.1->Topic=testTopic->action=write" |
Create testTopic
.
$ kafka-topics.sh --create --zookeeper localhost:2181 --replication-factor 1 --partitions 1 --topic testTopic $ kafka-topics.sh --list --zookeeper localhost:2181 testTopic |
Produce to testTopic
. Note that you will have to pass a config file, producer.properties, with information on jaas conf and other kerberos authentication related information. Here is more information.
$ kafka-console-producer.sh --broker-list localhost:9092 --topic testTopic --producer.config producer.properties This is a message This is another message |
Allow users in test-group
to describe a consumer group, testconsumergroup
, that it will be starting or joining.
$ kafka-sentry -gpr -r test -p "Host=127.0.0.1->Consumergroup=testconsumergroup->action=describe" |
Allow users in test-group
to read from a consumer group, testconsumergroup
, that it will be starting or joining.
$ kafka-sentry -gpr -r test -p "Host=127.0.0.1->Topic=testTopic->action=read" |
Allow users in test-group
to read from testTopic
from localhost, this will allow the users to consumer from testTopic
.
$ kafka-sentry -gpr -r test -p "Host=127.0.0.1->Topic=testTopic->action=read" |
Consume from testTopic
. Note that you will have to pass a config file, consumer.properties, with information on jaas conf and other kerberos authentication related information. The config file must also specify group.id
as testconsumergroup
. Here is more information.
$ kafka-console-consumer.sh --zookeeper localhost:2181 --topic testTopic --from-beginning --consumer.config consumer.properties This is a message This is another message |
Following numbers are obtained by running Kafka's ProducerPerformance tests on a three node Kafka cluster, one node Zookeeper and one node Sentry service. All tests are run from same host and test topic has three partitions with three set as replication factor.
Though Kafka authorization via Sentry works and it works nicely, there are a few more things that can be done here.