THIS IS A TEST INSTANCE. ALL YOUR CHANGES WILL BE LOST!!!!
Export enrichment node to the environment variable
Export zookeeper url
Export kafka broker url
Reference ambari url
0.xBETA
In this blog post we will walk through what it takes to setup a new telemetry source in Metron. For this example we will setup a new sensor, capture the sensor logs, pipe the logs to Kafka, pick up the logs with a Metron parsing topology, parse them, and run them through the Metron stream processing pipeline.
Our example sensor will be a Squid Proxy. Squid is a caching proxy for the Web supporting HTTP, HTTPS, FTP, and more. Squid logs are simple to explain and easy to parse and the velocity of traffic coming from Squid is representative of a a typical network-based sensor. Hence, we feel it's a good telemetry to use for this tutorial.
Step 1:
...
Download the Metron
...
source code
If you are not running Metron from the USB stick you need to download and build the code. Please Prior to going through this tutorial make sure you have Metron properly built and tested. Please see here for full Metron installation and validation instructions. Verify that the project has been built before creating the VM. First lets get Metron from Apache.
git clone https://git-wip-us.apache.org/repos/asf/incubator-metron.gitgithub.com/apache/metron.git
git tag -l
Now you will see a list of Metron releases. You will see major releases, minor releases, and release candidaescandidates. Refer to the Metron website with regards to which is the current stable release recommended for downloading. Once you select the Metron release run the following comand command to download it:
cdincubator-metron
git checkout tags/[MetronReleaseVersion]
Step 2: Build the Metron dev environment
Now that we have downloaded Metron and checked out the desired version, we need to build it. For the purposes of this exercise we will build without running through Metron's unit and integration test suites. To do so run the following command:
mvn clean package
Now we have downloaded and built metron it's on to the next step. Next we need to make a decision about the Metron environment and which parts of Metron we would like to build. For the purpose of this exercise we will assume that we are building full metron on the QuickDev image or Amazon AWS. We will provide instructions for both.
Step 2a : Setup Setup the QuickDev Image
If you want to take Metron for a spin local on your laptop you need to setup the QuickDev environment. The QuickDev environment is primarily intended for developers and people who want to spin up Metron quickly without incurring costs on AWS. But fair warning, this environment is not meant for production and is not performant. It is is intended merely for demonstration and development. To spin up QuickDev perform the following steps:
vagrant plugin install vagrant-hostmanager
cd metron-deployment/vagrant/quick-dev-platform
./launch_dev_image.sh
vagrant ssh
After executing the above commands a Metron VM will be built (called node1) and you will be logged in as user vagrant. There will be 4 topologies running but one must be stopped because the VM only has 4 Storm worker slots available.
Step 2b : Setup a Full AWS Metron Environment
The AWS environment is intended to install Metron in the AWS cloud. By default Metron is installed with a few demo sensors. To build the AWS environment in full run the following commands:
cd metron-deployment/amazon-ec2/
./run.sh
This will spin up your full AWS environment. At the end of the install the environment summary will be displayed as follows:
ok: [localhost] => {
"Success": [
"Apache Metron deployed successfully",
" Metron @ [METRON_HOST]:5000",
" Ambari @ [AMBARI_HOST]:8080",
" Sensor Status @ [SENSOR_HOST]:2812",
" Topology Status @ [MONIT_HOST]:2812",
"For additional information, see https://metron.incubator.apache.org/'"
]
}
Step 2c : Setup a Partial AWS Metron Environment
If you don't want to setup a full Metron environment you can deploy individual Metron modules. To do so, you need to define the Metron inventory. A sample inventory is provided with Metron to make custom inventories easier to define. To get to the inventory run the following command:
cd incubator-metron/metron-deployment/inventory/metron_example
There you will see two files: hosts and environment_vars/all. The first thing we need to define are environment variables for the Ansible scripts in the environment_vars/all file. This file contains granular settings for installation of each Metron component, including enabling or disabling the installation of a specific component as well as additional specifications. The second file you need to define is the hosts file. The hosts file defines the Metron cluster and what role individual nodes in the cluster will play. The following roles are possible:
- [ambari_master] - host running Ambari
- [ambari_slaves] - all Ambari-managed hosts
- [metron_kafka_topics] - host used to create the Kafka topics required by Metron. Requires a Kafka broker.
- [meron_hbase_tables] - host used to create the HBase tables required by Metron. Requires a HBase client.
- [enrichment] - submits the topology code to Storm and requires a Storm client
- [search] - host(s) where Elasticsearch will be installed
- [web] - host where the Metron UI and underlying services will be installed
- [sensors] - host where network data will be collected and published to Kafka
Once you configure the hosts and services, then run the following command:
cd incubator-metron/metron-deployment/playbooks
ansible-playbook -i ../inventory/project_name metron_install.yml --skip-tags="solr"
Step 2d : Setup Metron on an existing Ambari-managed Cluster (bare metal or AWS)
For this part it does not matter if you are installing core Metron components on bare metal or VMs. However it does matter for Metron sensors, as they need to be custom-compiled to the specific environment on which they are running. Currently we only support sensor installs on CentOS 6.7, Ansible 2.0.0.2, Java 8, and Intel x520 series of network cards.
First, we pre-assume that the Ambari cluster already exists. If it does not exist, you can deploy it by using the following set of instructions:
https://ambari.apache.org/1.2.1/installing-hadoop-using-ambari/content/ambari-chap1.html
The sample configuration for a 12-node cluster would be as follows:
node1 - [ambari_master]
node2 - [ambari_slaves]
node3 - [ambari_slaves]
node4 - [ambari_slaves]
node5 - [ambari_slaves]
node6 - [ambari_slaves]
node7 - [ambari_slaves]
node8 - [ambari_slaves]
node9-12 provision the OS + Java, but leave alone for now
Now we need to define a hosts file to install Metron on top of this cluster in the inventory hosts file. First we need to define a host to provision Metron Hbase tables (if using canned enrichments provided with Metron)
[metron_hbase_tables]
node9
Then we need to define a host to provision Metron's Kafka topics (if using canned sensors provided with Metron)
[metron_kafka_topics]
node9
The setup the node for your PCAP server/replay capability (if using the canned PCAP probe provided with Metron)
[pcap_server]
node9
Then define the node which will contain Storm jars and deployment scripts for Metron's parser and enrichment telemetries
#3rd ambari_slave
[enrichment]
node1
Then define the nodes which will contain Elastic Search master and slave nodes
#1 or more
[search]
node10
node11
node12
Then define nodes which will contain canned Metron sensors YAF, Bro, Snort, PCAP (if using caneed Metron sensors)
#1 only
[sensors]
node1
Then define the node where Kibana will be installed
#same as mysql in 12 node topology
[web]
node12
#same as mysql in 12 node topology
[web]
node12
Finally define the node where MySQL will be installed (if using Geo enrichment)
[mysql]
node12
Then based on your cluster definition edit group_vars/all file and then run:
ansible-playbook -i ../inventory/project_name metron_install.yml --skip-tags="solr"
This will automatically install Metron on an Ambari-managed cluster. For more detailed instructions please refer to:
https://github.com/dlyle65535/incubator-metron/blob/METRON-260/metron-deployment/README.md
Step 3 : Installing a sample sensor
Log into the sensors node and install the squid sensor. If you are on the QuickDev platform your VM will be called node1. If you are on AWS environment your sensor node will be tagged with the [sensors] tag. You can look through the AWS console to find which node in your cluster has this tag. Once you log into this node you can install the Squid sensor.
sudo yum install squid
sudo service squid start
This will run through the install and the Squid sensor will be installed and started. Now lets look at Squid logs.
sudo su -
cd /var/log/squid
ls
You see that there are three types of logs available: access.log, cache.log, and squid.out. We are interested in access.log as that is the log that records the proxy usage. We see that initially the log is empty. Lets generate a few entries for the log.
RUN more complex data
squidclient -g 20 http://www.aliexpress.com/af/shoes.html?ltype=wholesale&d=y&origin=n&isViewCP=y&catId=0&initiative_id=SB_20160622082445&SearchText=shoes
squidclient -g 20 "http://www.help.1and1.co.uk/domains-c40986/transfer-domains-c79878"
squidclient http://www.cnn.com
squidclient http://www.nba.com
vi /var/log/squid/access.log
In production environments you would configure your users web browsers to point to the proxy server, but for the sake of simplicity of this tutorial we will use the client that is packaged with the Squid installation After we use the client to simulate proxy requests the Squid log entries would look as follows:
setup our environment. There are a few choices as described here https://github.com/apache/metron/tree/master/metron-deployment. We'll choose Centos 6 for this example.
cd metron/metron-deployment/development/centos6
vagrant up
This will build Metron (without running the tests), package up relevant project artifacts as RPMs, setup and install Ambari to install and manage the single-node Hadoop cluster, and finally install Metron. Once the Vagrant command is finished, you should have a fully-running and self-contained virtual environment with Metron running inside of it.
TASK [deployment-report : debug] ***********************************************
ok: [node1] => {
"success": [
"Apache Metron deployed successfully",
" Ambari @ http://node1:8080",
" Zookeeper @ node1:2181",
" Kafka @ node1:6667",
"For additional information, see https://metron.apache.org/'"
]
}
PLAY RECAP *********************************************************************
node1 : ok=152 changed=64 unreachable=0 failed=0
Step 3 : Installing a sample sensor
Log into the sensors node and install the squid sensor. If you are on the local FullDev Vagrant development platform your VM will be called node1. See https://github.com/apache/metron/tree/master/metron-deployment/development/centos6 for example. If you are on AWS environment your sensor node will be tagged with the [sensors] tag. You can look through the AWS console to find which node in your cluster has this tag. For the Centos 6 local development environment, login as follows with password (in all lowercase) "vagrant"
ssh root@node1
Once you log into the sensor node you can install the Squid sensor.
sudo yum install squid
sudo service squid start
This will run through the install and the Squid sensor will be installed and started. Now let's look at Squid logs.
sudo su -
cd /var/log/squid
ls
You see that there are three types of logs available:
- access.log
- cache.log
- squid.out
We are interested in access.log as that is the log that records the proxy usage. We see that initially the log is empty. Lets generate a few entries for the log.
squidclient "http://www.aliexpress.com/af/shoes.html?ltype=wholesale&d=y&origin=n&isViewCP=y&catId=0&initiative_id=SB_20160622082445&SearchText=shoes"squidclient "http://www.help.1and1.co.uk/domains-c40986/transfer-domains-c79878"squidclient "http://www.pravda.ru/science/"squidclient "https://www.google.com/maps/place/Waterford,+WI/@42.7639877,-88.2867248,12z/data=!4m5!3m4!1s0x88059e67de9a3861:0x2d24f51aad34c80b!8m2!3d42.7630722!4d-88.2142563"squidclient "http://www.brightsideofthesun.com/2016/6/25/12027078/anatomy-of-a-deal-phoenix-suns-pick-bender-chriss"squidclient "https://www.microsoftstore.com/store/msusa/en_US/pdp/Microsoft-Band-2-Charging-Stand/productID.329506400"squidclient "http://www.autonews.com/article/20151115/RETAIL04/311169971/toyota-fj-cruiser-is-scarce-hot-and-high-priced"squidclient "https://tfl.gov.uk/plan-a-journey/"squidclient "https://www.facebook.com/Africa-Bike-Week-1550200608567001/"squidclient "http://www.ebay.com/itm/02-Infiniti-QX4-Rear-spoiler-Air-deflector-Nissan-Pathfinder-/172240020293?fits=Make%3AInfiniti%7CModel%3AQX4&hash=item281a4e2345:g:iMkAAOSwoBtW4Iwx&vxp=mtr"squidclient "http://www.recruit.jp/corporate/english/company/index.html"squidclient "http://www.lada.ru/en/cars/4x4/3dv/about.html"squidclient "http://www.help.1and1.co.uk/domains-c40986/transfer-domains-c79878"squidclient "http://www.aliexpress.com/af/shoes.html?ltype=wholesale&d=y&origin=n&isViewCP=y&catId=0&initiative_id=SB_20160622082445&SearchText=shoes"
In production environments you would configure your users web browsers to point to the proxy server, but for the sake of simplicity of this tutorial we will use the client that is packaged with the Squid installation After we use the client to simulate proxy requests the Squid log entries would look as follows:
1461576382.642 161
1467011157.401 415 127.0.0.1 TCP_MISS/200 337891 GET http://www.aliexpress.com/af/shoes.html? - DIRECT/207.109.73.154 text/html1467011158.083 671 127.0.0.1 TCP_MISS/200 41846 GET http://www.help.1and1.co.uk/domains-c40986/transfer-domains-c79878 - DIRECT/212.227.34.3 text/html1467011159.978 1893 127.0.0.1 TCP_MISS/200 153925 GET http://www.pravda.ru/science/ - DIRECT/185.103.135.90 text/html1467011160.044 58 127.0.0.1 TCP_MISS/302 1471 GET https://www.google.com/maps/place/Waterford,+WI/@42.7639877,-88.2867248,12z/data=cdcd/var/log/squidm5squidclienthttp://www.aliexpress.com/af/shoes.html? - DIRECT/172.217.3.164 text/html1467011160.145 155 127.0.0.1 TCP_MISS/200 133234 GET http://www.brightsideofthesun.com/2016/6/25/12027078/anatomy-of-a-deal-phoenix-suns-pick-bender-chriss - DIRECT/151.101.41.52 text/html1467011161.224 1073 127.0.0.1 TCP_MISS/200 141323 GET https://www.microsoftstore.com/store/msusa/en_US/pdp/Microsoft-Band-2-Charging-Stand/productID.329506400 - DIRECT/2.19.142.162 text/html1467011161.491 262 127.0.0.1 TCP_MISS/302 1955 GET http://www.autonews.com/article/20151115/RETAIL04/311169971/toyota-fj-cruiser-is-scarce-hot-and-high-priced - DIRECT/54.88.37.253 text/html1467011162.627 1133 127.0.0.1 TCP_MISS/200 88544 GET https://tfl.gov.uk/plan-a-journey/ - DIRECT/54.171.145.187 text/html1467011163.515 879 127.0.0.1 TCP_MISS/200 461930 GET https://www.facebook.com/Africa-Bike-Week-1550200608567001/ - DIRECT/69.171.230.68 text/html1467011164.286 749 127.0.0.1 TCP_MISS/200 190407 GET http://www.ebay.com/itm/02-Infiniti-QX4-Rear-spoiler-Air-deflector-Nissan-Pathfinder-/172240020293? - DIRECT/23.74.62.44 text/html1467011164.447 128200 103701
127.0.0.1 TCP_MISS/cnn.com/
404 12920 GET http://www.199
recruit.jp/corporate/english/company/index.html - DIRECT/27
23.79
74.73
66.1461576442
205 text/html228
1467011166.159
125137183
1659 127.0.0.1 TCP_MISS/200nba.com/
69469 GET http://www.66
lada.ru/en/cars/4x4/3dv/about.html - DIRECT/210
195.41
144.9
198.
77 text/html
The format of the log is timestamp | time elapsed | remotehost | code/status | bytes | method | URL rfc931 peerstatus/peerhost | type
Now that we have the sensor set up and generating logs we need to figure out how to pipe these logs to a Kafka topic. To do so the first thing we need to do is setup a new Kafka topic for Squid.
Change to environment variable for zookeeper
/usr/hdp/current/kafka-broker/bin//kafka-topics.sh --zookeeper localhost:2181 --create --topic squid --partitions 1 --replication-factor 1
/usr/hdp/current/kafka-broker/bin//kafka-topics.sh --zookeeper localhost:2181 --list
The following commands will setup a new Kafka topic for squid. Now let's test how we can pipe the Squid logs to Kakfka
tail /var/log/squid/access.log | /usr/hdp/current/kafka-broker/bin/kafka-console-producer.sh --broker-list node1:6667 --topic squid
./kafka-console-consumer.sh --zookeeper node1:2181 --topic squid --from-beginning
This should ingest our Squid logs into Kafka. Now we are ready to tackle the Metron parsing topology setup. The first thing we need to do is decide if we will be using the Java-based parser of a Grok-based parser for the new telemetry. In this example we will be using the Grok parser. Grok parser is perfect for structured or semi-structured logs that are well understood (check) and telemetries with lower volumes of traffic (check). The first thing we need to do is define the Grok expression for our log. Refer to Grok documentation for additional details. In our case the pattern is:
SQUID_DELIMITED %{NUMBER:timestamp} %{SPACE:UNWANTED} %{INT:elapsed} %{IPV4:ip_src_addr} %{WORD:action}/%{NUMBER:code} %{NUMBER:bytes} %{WORD:method} %{NOTSPACE:url} - %{WORD:UNWANTED}\/%{IPV4:ip_dst_addr} %{WORD:UNWANTED}\/%{WORD:UNWANTED}
- this is already pre-loaded under /apps/metron/patterns/squid
Notice that I apply the UNWANTED tag for any part of the message that I don't want included in my resulting JSON structure. Finally, notice that I applied the naming convention to the IPV4 field by referencing the following list of field conventions. The last thing I need to do is to validate my Grok pattern to make sure it's valid. For our test we will be using a free Grok validator called Grok Constructor. A validated Grok expression should look like this:
update graphic
Now that the Grok pattern has been defined we need to save it and move it to HDFS. Existing Grok parsers that ship with Metron are staged under /apps/metron/patterns/
don't need to do this step if patterns already pre-loaded
[root@node1 bin]# hdfs dfs -ls /apps/metron/patterns/
Found 5 items
-rw-r--r-- 3 hdfs hadoop 13427 2016-04-25 07:07 /apps/metron/patterns/asa
-rw-r--r-- 3 hdfs hadoop 5203 2016-04-25 07:07 /apps/metron/patterns/common
-rw-r--r-- 3 hdfs hadoop 524 2016-04-25 07:07 /apps/metron/patterns/fireeye
-rw-r--r-- 3 hdfs hadoop 2552 2016-04-25 07:07 /apps/metron/patterns/sourcefire
-rw-r--r-- 3 hdfs hadoop 879 2016-04-25 07:07 /apps/metron/patterns/yaf
We need to move our new Squid pattern into the same directory. Create a file from the grok pattern above:
touch /tmp/squid
vi /tmp/squid
Then move it to HDFS:
su - hdfs
hdfs dfs -put /tmp/squid /apps/metron/patterns/
exit
Now that the Grok pattern is staged in HDFS we need to define a parser configuration for the Metron Parsing Topology. The configurations are kept in Zookeeper so the sensor configuration must be uploaded there after it has been created. A Grok parser configuration follows this format:
{
"parserClassName": "org.apache.metron.parsers.GrokParser",
"sensorTopic": "sensor name",
"parserConfig": {
"grokPath": "grok pattern",
"patternLabel": "grok label",
... other optional fields
}
}
Create a Squid Grok parser configuration file at /usr/metron/0.1BETA/config/zookeeper/parsers/squid.json with the following contents:
reference stellar docs
relink
{
"parserClassName": "org.apache.metron.parsers.GrokParser",
"sensorTopic": "squid",
"parserConfig": {
"grokPath": "/apps/metron/patterns/squid",
"patternLabel": "SQUID_DELIMITED",
"timestampField": "timestamp"
},
"fieldTransformations" : [
{
"transformation" : "MTL"
,"output" : [ "full_hostname", "domain_without_subdomains" ]
,"config" : {
"full_hostname" : "URL_TO_HOST(url)"
,"domain_without_subdomains" : "DOMAIN_REMOVE_SUBDOMAINS(full_hostname)"
}
}
]
1467011166.543 401 127.0.0.1 TCP_MISS/200 41846 GET http://www.help.1and1.co.uk/domains-c40986/transfer-domains-c79878 - DIRECT/212.227.34.3 text/html1467011168.519 445 127.0.0.1 TCP_MISS/200 336155 GET http://www.aliexpress.com/af/shoes.html? - DIRECT/207.109.73.154 text/html
The format of the log is:
timestamp | time elapsed | remotehost | code/status | bytes | method | URL rfc931 peerstatus/peerhost | type
Now that we have the sensor set up and generating logs we need to figure out how to pipe these logs to a Kafka topic. To do so the first thing we need to do is setup a new Kafka topic for Squid.
Step 4 : Define Environment Variables
If you are using the quick-dev image your links are:
Ambari: http://node1:8080/
Storm: http://node1:8744/index.html
Now lets setup the following environment variables on node1 to make it easier to navigate and carry over the commands from full-dev to AWS or bare metal deployment.
source /etc/default/metron
export HDP_HOME="/usr/hdp/current"
Note: It's worth checking the the values of ZOOKEEPER and BROKERLIST before continuing. You should supply a comma-delimited list of host:port items for the ZOOKEEPER and BROKERLIST variables if you are running in an environment with multiple hosts for Zookeeper and the Kafka brokers.
Step 5 : Create Kafka topics and ingest sample data
${HDP_HOME}/kafka-broker/bin/kafka-topics.sh --zookeeper $ZOOKEEPER --create --topic squid --partitions 1 --replication-factor 1
${HDP_HOME}/kafka-broker/bin/kafka-topics.sh --zookeeper $ZOOKEEPER --list
The following commands will setup a new Kafka topic for squid. Now let's test how we can pipe the Squid logs to Kakfka
cat /var/log/squid/access.log | ${HDP_HOME}/kafka-broker/bin/kafka-console-producer.sh --broker-list $BROKERLIST --topic squid
${HDP_HOME}/kafka-broker/bin/kafka-console-consumer.sh --bootstrap-server $BROKERLIST --topic squid --from-beginning
Note: The following steps for manually creating the Grok expression, copying the pattern to HDFS, and creating the parser and indexing json configs for the sensor is no longer necessary in full dev. The files are installed by default and you can simply start the squid topology as described below to achieve the end result of these steps.
This should ingest our Squid logs into Kafka. Now we are ready to tackle the Metron parsing topology setup. The first thing we need to do is decide if we will be using the Java-based parser of a Grok-based parser for the new telemetry. In this example we will be using the Grok parser. Grok parser is perfect for structured or semi-structured logs that are well understood (check) and telemetries with lower volumes of traffic (check). The first thing we need to do is define the Grok expression for our log. Refer to Grok documentation for additional details. In our case the pattern is:
SQUID_DELIMITED %{NUMBER:timestamp}[^0-9]*%{INT:elapsed} %{IP:ip_src_addr} %{WORD:action}/%{NUMBER:code} %{NUMBER:bytes} %{WORD:method} %{NOTSPACE:url}[^0-9]*(%{IP:ip_dst_addr})?
Notice that I apply the UNWANTED tag for any part of the message that I don't want included in my resulting JSON structure. Finally, notice that I applied the naming convention to the IPV4 field by referencing the following list of field conventions. The last thing I need to do is to validate my Grok pattern to make sure it's valid. For our test we will be using a free Grok validator called Grok Constructor. A validated Grok expression should look like this:
Now that the Grok pattern has been defined we need to save it and move it to HDFS. Existing Grok parsers that ship with Metron are staged under /apps/metron/patterns/
First we do a directory listing to see which patterns are available with the platform
# hdfs dfs -ls /apps/metron/patterns/
Found 7 items
-rwxr-xr-x 1 metron hdfs 13748 2019-08-21 20:37 /apps/metron/patterns/asa
-rwxr-xr-x 1 metron hdfs 5202 2019-08-21 20:37 /apps/metron/patterns/common
-rwxr-xr-x 1 metron hdfs 524 2019-08-21 20:37 /apps/metron/patterns/fireeye
-rwxr-xr-x 1 metron hdfs 2551 2019-08-21 20:37 /apps/metron/patterns/sourcefire
-rwxr-xr-x 1 metron hdfs 180 2019-08-21 20:37 /apps/metron/patterns/squid
-rwxr-xr-x 1 metron hdfs 2220 2019-08-21 20:37 /apps/metron/patterns/websphere
-rwxr-xr-x 1 metron hdfs 879 2019-08-21 20:37 /apps/metron/patterns/yaf
Now we add a new pattern need to move our new Squid pattern into the same directory. Create a file from the grok pattern above:
touch /tmp/squid
vi /tmp/squid
Then move it to HDFS:
su - hdfs
**if the pattern already exists and you need to replace also run hdfs dfs -rm /apps/metron/patterns/squid
hdfs dfs -put /tmp/squid /apps/metron/patterns/
exit
Now that the Grok pattern is staged in HDFS we need to define a parser configuration for the Metron Parsing Topology. The configurations are kept in Zookeeper so the sensor configuration must be uploaded there after it has been created. A Grok parser configuration follows this format:
{
"parserClassName": "org.apache.metron.parsers.GrokParser",
"sensorTopic": "sensor name",
"parserConfig": {
"grokPath": "grok pattern",
"patternLabel": "grok label",
... other optional fields
}
}
There is a pre-packaged Squid Grok parser configuration file at ${METRON_HOME}/config/zookeeper/parsers/squid.json with the following contents:
{
"parserClassName": "org.apache.metron.parsers.GrokParser",
"sensorTopic": "squid",
"parserConfig": {
"grokPath": "/patterns/squid",
"patternLabel": "SQUID_DELIMITED",
"timestampField": "timestamp"
},"fieldTransformations" : [
{
"transformation" : "STELLAR"
,"output" : [ "full_hostname", "domain_without_subdomains" ]
,"config" : {
"full_hostname" : "URL_TO_HOST(url)"
,"domain_without_subdomains" : "DOMAIN_REMOVE_SUBDOMAINS(full_hostname)"
}
}
]}
Notice the use of the fieldTransformations in the parser configuration. Our Grok Parser is set up to extract the URL, but really we want just the domain or even the domain without subdomains. To do this, we can use the Metron Transformation Language field transformation. The Metron Transformation Language is a Domain Specific Language which allows users to define extra transformations to be done on the messages flowing through the topology. It supports a wide range of common network and string related functions as well as function composition and list operations. In our case, we extract the hostname from the URL via the URL_TO_HOST function and remove the domain names with DOMAIN_REMOVE_SUBDOMAINS thereby creating two new fields, "full_hostname" and "domain_without_subdomains" to each message.
We can also setup index types and batch sizing. Add the following lines to a file named ${METRON_HOME}/config/zookeeper/indexing/squid.json
{"hdfs" : {"index": "squid","batchSize": 5,"enabled" : true},"elasticsearch" : {"index": "squid","batchSize": 5,"enabled" : true},"solr" : {"index": "squid","batchSize": 5,"enabled" : true}}
Another thing we can do is validate our messages. Lets say we wanted to make sure that source IPs and destination IPs are valid. The validators are global so we set them up on the global JSON and push them into Zookeeper. To do so perform the following commands:
vi ${METRON_HOME}/config/zookeeper/global.json
and update the json to contain at least the following:
{
"es.clustername": "metron",
"es.ip": "node1:9300",
"es.date.format": "yyyy.MM.dd.HH","parser.error.topic": "indexing",
"fieldValidations" : [
{
"input" : [ "ip_src_addr", "ip_dst_addr" ],
"validation" : "IP",
"config" : {
"type" : "IPV4"
}
}
]}
A script is provided to upload configurations to Zookeeper. Upload the configs with the PUSH option.
${METRON_HOME}/bin/zk_load_configs.sh -i ${METRON_HOME}/config/zookeeper -m PUSH -z $ZOOKEEPER
And we can verify our configurations have been uploaded by using the DUMP command.
${METRON_HOME}/bin/zk_load_configs.sh -m DUMP -z $ZOOKEEPER
Now, install an Elasticsearch template for your new sensor so that we can effectively query results in the Metron Alerts UI.
Note: This is a new step that is necessary as of the meta alerts feature and Elasticsearch 5.6.2 upgrade.
Run the following commands from the CLI.
curl -XPUT 'http://node1:9200/_template/squid_index' -d '{"template": "squid_index*","mappings": {"squid_doc": {"dynamic_templates": [{"geo_location_point": {"match": "enrichments:geo:*:location_point","match_mapping_type": "*","mapping": {"type": "geo_point"}}},{"geo_country": {"match": "enrichments:geo:*:country","match_mapping_type": "*","mapping": {"type": "keyword"}}},{"geo_city": {"match": "enrichments:geo:*:city","match_mapping_type": "*","mapping": {"type": "keyword"}}},{"geo_location_id": {"match": "enrichments:geo:*:locID","match_mapping_type": "*","mapping": {"type": "keyword"}}},{"geo_dma_code": {"match": "enrichments:geo:*:dmaCode","match_mapping_type": "*","mapping": {"type": "keyword"}}},{"geo_postal_code": {"match": "enrichments:geo:*:postalCode","match_mapping_type": "*","mapping": {"type": "keyword"}}},{"geo_latitude": {"match": "enrichments:geo:*:latitude","match_mapping_type": "*","mapping": {"type": "float"}}},{"geo_longitude": {"match": "enrichments:geo:*:longitude","match_mapping_type": "*","mapping": {"type": "float"}}},{"timestamps": {"match": "*:ts","match_mapping_type": "*","mapping": {"type": "date","format": "epoch_millis"}}},{"threat_triage_score": {"mapping": {"type": "float"},"match": "threat:triage:*score","match_mapping_type": "*"}},{"threat_triage_reason": {"mapping": {"type": "text","fielddata": "true"},"match": "threat:triage:rules:*:reason","match_mapping_type": "*"}},{"threat_triage_name": {"mapping": {"type": "text","fielddata": "true"},"match": "threat:triage:rules:*:name","match_mapping_type": "*"}}],"properties": {"timestamp": {"type": "date","format": "epoch_millis"},"source:type": {"type": "keyword"},"ip_dst_addr": {"type": "ip"},"ip_dst_port": {"type": "integer"},"ip_src_addr": {"type": "ip"},"ip_src_port": {"type": "integer"},"alert": {"type": "nested"},"metron_alert" : {
"type" : "nested"
},"guid": {"type": "keyword"}}}}}'# Verify the template installs as expectedcurl -XGET 'http://node1:9200/_template/squid_index?pretty'
This template accomplishes two things:
- Sets up default mappings for metron-specific types, e.g. timestamps.
- Sets up types for properties that will come from the parsed data, e.g. ip_src_addr.
If you're using the Full dev environment, you might want to stop some of the other parsers to free up resources.
for parser in bro__snort__yaf profiler pcap batch_indexing; do storm kill parser; done
Now start the new squid parser topology:
${METRON_HOME}/
...
Notice the use of the fieldTransformations in the parser configuration. Our Grok Parser is set up to extract the URL, but really we want just the domain or even the domain without subdomains. To do this, we can use the Metron Transformation Language field transformation. The Metron Transformation Language is a Domain Specific Language which allows users to define extra transformations to be done on the messages flowing through the topology. It supports a wide range of common network and string related functions as well as function composition and list operations. In our case, we extract the hostname from the URL via the URL_TO_HOST function and remove the domain names with DOMAIN_REMOVE_SUBDOMAINS thereby creating two new fields, "full_hostname" and "domain_without_subdomains" to each message.
A script is provided to upload configurations to Zookeeper. Upload the new parser config to Zookeeper:
/usr/metron/0.1BETA/bin/zk_load_configs.sh --mode PUSH -i /usr/metron/0.1BETA/config/zookeeper -z node1:2181
Start the new squid parser topology:
/usr/metron/0.1BETA/bin/start_parser_topology.sh -k node1:6667 $BROKERLIST -z node1:2181 $ZOOKEEPER -s squid
Navigate to the squid parser topology in the Storm UI at http://node1:8744/index.html and verify the topology is up with no errors:
Update Graphic
CREATE ES template before deployment
Now that we have a new running squid parser topology, generate some data to parse by running this command several times:
sudo tail /var/log/squid/access.log |/usr/hdp/current${HDP_HOME}/kafka-broker/bin/kafka-console-producer.sh --broker-listnode1:6667$BROKERLIST --topic squid
Refresh the Storm UI and it should report data being parsed:
...
yellow open yaf_index_2016.04.25.15 5 1 5485 0 4mb 4mb yellow open snort_index_2016.04.26.12 5 1 244525485 0 14.4mb 14. 4mb
yellow open brosnort_index_2016.04.2526.16 12 5 1 129524452 0 114.9mb4mb 14.4mb 1.9mb
yellow open squidbro_index_2016.04.2625.1316 5 1 1295 1 0 71.3kb9mb 71.3kb 9mb
yellow open yafsquid_index_2016.04.25.17 5 1 26.13 307505 1 0 1 17.4mb 17.4mb
In order to verify that the messages were indexed correctly first install elastic search Head plugin:
/usr/share/elasticsearch/bin/plugin -install mobz/elasticsearch-head/1.x
And navigate to http://node1:9200/_plugin/head/
There you will see parsed message + performance timestamps. We will discuss the performance timestamps in another blog entry.
...
0 7.3kb 7.3kb yellow open yaf_index_2016.04.25.17 5 1 30750 0 17.4mb 17.4mb In order to verify that the messages were indexed correctly first install elastic search Head plugin:
| Note |
|---|
The Elasticsearch Head plugin is no longer available post 5.x. You have 3 options now:
|
And navigate to http://node1:9200/_plugin/head/ one of the above mentioned tools for data exploration.
There you will see parsed message + performance timestamps. We will discuss the performance timestamps in another blog entry.
Now lets see how we create a Kibana dashboard to visualize data in metron. First click on Visualize, select a squid index, and add the fields you wan to display
Then click on save to save the query and import it into the main Metron dashboard:
By convention the index where the new messages will be indexed is called squid_index_[timestamp] and the document type is squid_doc.
Now that we have the messages parsed and indexed we need to setup a Kibana dashboard. To do so access the dashboard on http://node1:5000/#/dashboard/file/default.json
To create a new ingest histogram we first need to setup a pinned query. Click on the query + button and pin a query for _type:squid_doc. This would look like:
Once the query is pinned it will show up in the pinned queries bar like so:
Once the query is established we can create a histogram panel. In the panel settings point the panel to listed to the Squid Logs pinned query you just created
And make sure that the time field points to the field called "timestamp:
Click OK and you should get a histogram that looks like this:
Now to add a detailed telemetry table create a new table panel, and similarly to the histogram panel point it to the Squid Logs pinned query. As a result the following table will be created:
