11 May 2015

Analyzing Weblogs with Elasticsearch in the Cloud, Part 1

By Christoffer Vig

UPDATE: This article refers to our hosted Elasticsearch offering by an older name, Found. Please note that Found is now known as Elastic Cloud.

Using Logstash and Kibana on Found by Elastic

This is part one of a two post blog series, aiming to demonstrate how to feed logs from IIS into Elasticsearch and Kibana via Logstash, using the hosted services provided by Found by Elastic. This post will deal with setting up the basic functionality and securing connections. Part 2 will show how to configure Logstash to read from IIS log files, and how to use Kibana 4 to visualize web traffic.

Getting the Bits

For this demo I will be running Logstash and Kibana from my Windows laptop. If you want to follow along, download and extract Logstash 1.5.RC4 or later, and Kibana 4.0.2 or later from https://www.elastic.co/downloads. We need Logstash 1.5, since it introduces support for SSL connections.

Creating an Elasticsearch Cluster

Creating a new trial cluster in Found is just a matter of logging in and pressing a button. It takes a few seconds until the cluster is ready, and a screen with some basic information on how to connect pops up. We need the address for the HTTPS endpoint, so copy that out.

Configuring Logstash

Now, with the brand new SSL connection option in Logstash, connecting to Found is as simple as this Logstash configuration:

input { stdin{} }
output {
  elasticsearch {
    protocol => http
    host => REPLACE_WITH_FOUND_CLUSTER_HOSTNAME (full uri in quotes)
    port => "9243" # Check the port also
    ssl => true
}
stdout { codec => rubydebug }
}

Save the file as found.conf

Start up Logstash using

bin\logstash.bat agent --verbose -f found.conf

You should see a message similar to

Create client to elasticsearch server on `https://....foundcluster.com:9243`: {:level=>:info}

Once you see “Logstash startup completed” type in your favorite test term on the terminal. Mine is “fisk” so I type that. You should see output on your screen showing what Logstash intends to pass on to Elasticsearch.

We want to make sure this actually hits the cloud, so open a browser window and paste the HTTPS link from before, append /_search to the URL and hit enter. You should now see the search results from your newly created Elasticsearch cluster, containing the favorite term you just typed in. We have a functioning connection from Logstash on our machine to Elasticsearch in the cloud! Congratulations!

Configuring Kibana 4

Kibana 4 comes with a built-in webserver. The configuration is done in a kibana.yml file in the config directory. Connecting to Elasticsearch in the cloud comes down to inserting the address of the Elasticsearch instance.

# The Elasticsearch instance to use for all your queries.
elasticsearch_url: `https://....foundcluster.com:9243`

Of course, we need to verify that this really works, so we open up Kibana on http://localhost:5601, select the Logstash index template, with the @timestamp data field as suggested, and open up the discover panel. Now, if there was less than 15 minutes since you inserted your favorite test term in Logstash (previous step), you should see it already. Otherwise, change the date range by clicking on the selector in the top right corner.

Kibana test

Kibana test

Locking it down

Found by Elastic has worked hard to make the previous steps easy. We created an Elasticsearch cluster, fed data into it and displayed in Kibana in less than 5 minutes. We must have forgotten something!? And yes, of course! Something about security. We made sure to use secure connections with SSL, and the address generated for our cluster contains a 32 character long, randomly generated list of characters, which is pretty hard to guess. Should, however, the address slip out of our hands, hackers could easily delete our entire cluster. And we don’t want that to happen. So let’s see how we can make everything work when we add some basic security measures.

Access Control Lists

Found by Elastic has support for access control lists, where you can set up lists of usernames and passwords, with lists of rules that deny/allow access to various paths within Elasticsearch. This makes it easy to create a “read only” user, for instance, by creating a user with a rule that only allows access to the /_search path. Found by Elastic has a sample configuration with users searchonly and readwrite. We will use these as starting point but first we need to figure out what Kibana needs.

Kibana 4 Security

Kibana 4 stores its configuration in a special index, by default named “.kibana”. The Kibana webserver needs write access to this index. In addition, all Kibana users need write access to this index, for storing dashboards, visualizations and searches, and read access to all the indices that it will query. More details about configuring Kibana 4 can be found in the article Using Kibana in a Production Environment, for details on the access demands of Kibana, see Configuring Roles for Kibana 4 Users.

For this demo, we will simply copy the “readwrite” user from the sample twice, naming one kibanaserver, the other kibanauser.

# Setting Access Control in Found: 
# Allow everything for the readwrite-user, kibanauser and kibanaserver
  - paths: ['.*']
    conditions:
      - basic_auth:
          users:
            - readwrite
            - kibanauser
            - kibanaserver
      - ssl:
          require: true
    action: allow

Press save and the changes are immediately effective. Try to reload the Kibana at http://localhost:5601, you should be denied access.

Open up the kibana.yml file from before and modify it:

# If your Elasticsearch is protected with basic auth, this is the user credentials
# used by the Kibana server to perform maintence on the kibana_index at statup. Your Kibana
# users will still need to authenticate with Elasticsearch (which is proxied thorugh
# the Kibana server)
kibana_elasticsearch_username: kibanaserver
kibana_elasticsearch_password: `KIBANASERVER_USER_PASSWORD`

Stop and start Kibana to effectuate settings. Now when Kibana starts up, you will be presented with a login box for HTTP authentication. Type in kibanauser as the username, and the password. You should now again be presented with the Discover screen, showing the previously entered favorite test term. Again, you may have to expand the time range to see your entry.

Logstash Security

Logstash will also need to supply credentials when connecting to Found by Elastic. We reuse permission from the readwrite user once again, this time giving the name “logstash”. It is simply a matter of supplying the username and password in the configuration file.

output {
    elasticsearch {
     …. 
    user => “logstash”,
    password => `LOGSTASH_USER_PASSWORD`
    }
}

Wrapping it up

This has been a short dive into Logstash and Kibana with Found by Elastic. The recent changes done in order to support the Shield plugin for Elasticsearch, Logstash and Kibana, make it very easy to use the secure features of Found by Elastic. In the next post we will look into feeding logs from IIS into Elasticsearch via Logstash, and visualizing the most used query terms in Kibana.