ELK stack introduction

Introduction to ELK stack
– 巨量資料處理、搜尋、及分析工具介紹 –
計資中心網路組邵喻美
madeline@ntu.edu.tw
1

Topics
• Why big data tool for network traffic and log analysis
• What is ELK stack, and why choose it
• ELK stack intro
• ELK use cases
• Implementation of ELK on network & account anomaly detection
2

Network operation and security management issues
• Lots of users
• Faculty & staff & students  more than 40000 users on campus
• Lots of systems
• Routers, firewalls, servers….
• Lots of logs
• Netflow, syslogs, access logs, service logs, audit logs….
• Nobody cares until something go wrong….
3

Logs & events analysis for network managements
• Logs & events collection from multiple sources
• Accept and parse different log formats
• Large amount, and various formats of data
• Scalable architecture
• Expert knowledge requirement
4

How we “traditional” system managers treat logs
• Set up one or more log servers for receiving logs from
servers/routers/appliances
• Unix commands -- grep + awk + sed + sort + uniq + perl + shell
script ….
• Cronjobs executed periodically
• compute stats and send out report/alert
• detect possible abnormal behavior and react accordingly
• Plain text reports or stats trends webpage
5

Amount of data….
• Router
• Netflow – 43GB daily
• Wifi
• NAT log – 4.8TB daily
• Auth log
• WAF/Firewall
• Server access logs & events
• Mail server log ~18GBdaily
• POP3 – avg. 7GB daily
• SMTP – avg. 1.75GB daily
• Exchange – avg. 140MB daily
• OWA – avg. 8.4GB daily
• MessageTrackingLog – avg. 100MB
daily
6

What is ELK, and why choose it
7

Splunk vs. ELK on Google Trend
One of the leaders in security
information and event
management (SIEM) market
How do Netflix, Facebook, Microsoft,
LinkedIn, and Cisco monitor their
logs? With ELK.
The ELK Stack is now
downloaded 500,000
times every month,
making it the world’s
most popular log
management platform
8

Why ELK?
• Rapid on-premise (or cloud) installation and easy to deploy
• Scales vertically and horizontally
• Easy and various APIs to use
• Ease of writing queries, a lot easier then writing a MapReduce job
• Availability of libraries for most programming/scripting languages
• Elastic offers a host of language clients for Elasticsearch, including Ruby, Python, PHP,
Perl, .NET, Java, and Javascript, and more
• Tools availability
• It’s free (open source), and it’s quick
9

Logstash
Data From Any Source
Elasticsearch
Instantly Search & Analyze
Kibana
Actionable Insight
Elasticsearch is a NoSQL
database that is based on the
Lucene search engine
 indexes and stores the
information
Kibana is a visualization
layer that works on top
of Elasticsearch
 presents the data in
visualizations that
provide actionable
insights
Logstash is a log pipeline
tool that accepts inputs
from various sources,
executes different
transformations, and
exports the data to
various targets
 collects and parses logs
10

ELK modules
Open Source —
• ElasticSearch
• Logstash
• Kibana
• Beats
• data shippers – collect, parse & ship
Extension plugins —
• Alerting (Watcher)
• Proactively monitoring and alerting based on
elasticsearch queries or conditions
• Security (Shield)
• Protect and provide security to elastic stack
• Monitoring (Marvel)
• Monitor and diagnose health and performance
of elastics cluster
• Graph
• discover and explore the relationships live in
data by adding relevance to your exploration
12

Connect Speedy Search with Big Data Analytics –
Elasticsearch for Apache Hadoop
ES-Hadoop -- a two-way connector
• Read and write data to ES and query it in real time
13

let’s look into ELK stack
14

Elasticsearch-Logstash-Kibana
16

Logstash architecture
Ip: 140.1.1.1
Ip: 140.1.1.1
City: Zurich
Country: CH
18

Logstash Input plugins
• Stdin – Reads events from standard input
• File – Streams events from files (similar to “tail -0F”)
• Syslog – Reads syslog messages as events
• Eventlog – Pulls events from theWindows Event Log
• Imap – read mail from an IMAP server
• Rss – captures the output of command line tools as an event
• Snmptrap – creates events based on SNMP trap messages
• Twitter – Reads events from the Twitter Streaming API
• Irc – reads events from an IRC server
• Exec – Captures the output of a shell command as an event
• Elasticsearch – Reads query results from an Elasticsearch cluster
• ….
20

Logstash Filter plugins
• grok – parses unstructured event data into fields
• Mutate – performs mutations on fields
• Geoip – adds geographical information about an IP address
• Date – parse dates from fields to use as the Logstash timestamp for
an event
• Cidr – checks IP addresses against a list of network blocks
• Drop – drops all events
• …
21

Logstash Output plugins
• Stdout – prints events to the standard output
• Csv – write events to disk in a delimited format
• Email – sends email to a specified address when output is received
• Elasticsearch – stores logs in Elasticsearch
• Exec – runs a command for a matching event
• File – writes events to files on disk
• mongoDB – writes events to MongoDB
• Redmine – creates tickets using the Redmine API
• ….
22

filter {
if [type] == "syslog" {
grok {
match => { "message" => "%{SYSLOGTIMESTAMP:syslog_timestamp}
%{SYSLOGHOST:syslog_hostname} %{DATA:syslog_program}(?:[%{POSINT:syslog_pid}])?:
%{GREEDYDATA:syslog_message}" }
add_field => [ "received_at", "%{@timestamp}" ]
add_field => [ "received_from", "%{host}" ]
}
date {
match => [ "syslog_timestamp", "MMM d HH:mm:ss", "MMM dd HH:mm:ss" ]
}
}
}
Dec 23 14:30:01 louis CRON[619]: (www-data) CMD (php /usr/share/cacti/site/poller.php
>/dev/null 2>/var/log/cacti/poller-error.log)
23

{
"message" => "Dec 23 14:30:01 louis CRON[619]: (www-data) CMD (php
/usr/share/cacti/site/poller.php >/dev/null 2>/var/log/cacti/poller-error.log)",
"@timestamp" => "2013-12-23T22:30:01.000Z",
"@version" => "1",
"type" => "syslog",
"host" => "0:0:0:0:0:0:0:1:52617",
"syslog_timestamp" => "Dec 23 14:30:01",
"syslog_hostname" => "louis",
"syslog_program" => "CRON",
"syslog_pid" => "619",
"syslog_message" => "(www-data) CMD (php /usr/share/cacti/site/poller.php
>/dev/null 2>/var/log/cacti/poller-error.log)",
"received_at" => "2013-12-23 22:49:22 UTC",
"received_from" => "0:0:0:0:0:0:0:1:52617",
"syslog_severity_code" => 5,
"syslog_facility_code" => 1,
"syslog_facility" => "user-level",
"syslog_severity" => "notice"
}
24

https://github.com/elastic/logstash/blob/v1.4.2/patterns/grok-patterns
27

Deploying and scaling Logstash
Minimal installation
Using Filters
28

Using log shipper to
minimize the resource
demands on Logstash
Scaling to a Larger
Elasticsearch Cluster
29

Managing Throughput
Spikes with Message
Queuing
30

Multiple Connections
for Logstash High
Availability
31

32

ElasticSearch
• Built on top of Apache Lucene™, a full-text search-engine library
• A Schema-free, REST & JSON based distributed search engine with real-time analytics
• Capable of scaling to hundreds of servers and petabytes of structured and unstructured data
• Open Source:Apache License 2.0
• Wikipedia uses Elasticsearch to provide full-text search with highlighted search snippets, andsearch-as-you-type and did-you-
mean suggestions
• The Guardian uses Elasticsearch to combine visitor logs with social-network data to provide real-time feedback to its
editors about the public’s response to new articles
• Stack Overflow combines full-text search with geolocation queries and uses more-like-this to find related questions and
answers
• GitHub uses Elasticsearch to query 130 billion lines of code
Real scalability comes from
horizontal scale
Schema-flexible
33

Elasticsearch vs. Relational DB
ElasticSearch Relational DB
Index Database
Type Table
Document Row
Field Column
Shard Partition
Mapping Schema
- (everything is indexed) Index
Query DSL (domain specific language) SQL
Shards are how
Elasticsearch distributes
data around your
cluster
34

What is a shard
• a shard is a single instance of Lucene, and is a complete search engine
in its own right
• Documents are stored and indexed in shards  shards are allocated
to nodes in your cluster
• As your cluster grows or shrinks, Elasticsearch will automatically
migrate shards between nodes so that the cluster remains balanced
• A shard can be either a primary shard or a replica shard
• Each document in your index belongs to a single primary shard
• A replica shard is just a copy of a primary shard
35

ElasticSearch clustering – single node cluster
• Node = running instance of ES
• Cluster = 1+ nodes with the same cluster.name
• Every cluster has 1 master node
• 1 Cluster can have any number of indexes
36

A cluster consists of one or more nodes with the same cluster.name
• All primary and replica shards are allocated
• Each index has one primary (P) and one replica (R) shard
• Clients talk to any node in the cluster
ElasticSearch clustering – adding a second node
37

ElasticSearch clustering – adding a third node
• More primary shards:
• faster indexing
• more scale
• More replicas:
• faster searching
• more failover
Increase the number of replicas
38

Creating, Indexing, and Deleting a document
1. The client sends a create, index, or delete request to Node 1
2. The node uses the document’s _id to determine that the document belongs to shard 0. It forwards
the request to Node 3, where the primary copy of shard 0 is currently allocated
3. Node 3 executes the request on the primary shard. If it is successful, it forwards the request in
parallel to the replica shards on Node 1 and Node 2. Once all of the replica shards report
success, Node 3 reports success to the coordinating node, which reports success to the client.
39

Retrieving a Document
1. The client sends a get request to node 1
2. The node uses the document’s _id to determine that the document belongs to shard
0. Copies of shard 0 exist on all three nodes. On this occasion, it forwards the request
to node 2.
3. Node 2 returns the document to node 1, which returns the document to the client.
For read requests, the coordinating
node will choose a different shard
copy on every request in order to
balance the load
40

Partial update to a document
When a primary shard forwards changes to its replica shards, it doesn’t
forward the update request. Instead it forwards the new version of the
full document.
41

Multidocument Patterns • the coordinating node knows in which
shard each document lives.
• It breaks up the multidocument request
into a multidocument request per shard,
and forwards these in parallel to each
participating node
• Once it receives answers from each
node, it collates their responses into a
single response
mget
bulk 42

Talking to Elasticsearch
• RESTful API with JSON over HTTP
• Over port 9200
• Access via web client, or command line by curl command
• JSON ( JavaScript Object Notation )  the standard format used by NoSQL
• Elasticsearch clients
• Java API, Java REST client, JavaScript API, PHP API, Python API, Perl API…
HTTP method or verb: GET,
POST, PUT, HEAD, or DELETE
43

Indexing a document
• Store a document in an index so that it can be retrieved and queried
• Like the INSERT keyword in SQL
44

Retrieving documents
• Using GET method to retrieve document
• We can retrieve a specific document if we happen to know its id
45

Performing Queries
• Using the q=<query> form performs a
full-text search by parsing the query
string value
• Query with query DSL, which is specified
using a JSON request body
46

Query DSL – Combining Filters
Bool Filter
47

Query DSL – Nesting Boolean Queries
48

49

Kibana
• Search, view, and interact with data stored in Elasticsearch indices
• Execute queries on data & visualize results in charts, tables, and maps
• Add/remove widgets
• Share/Save/Load dashboards
• Open Source:Apache License 2.0
50

Cisco Talos Security Intelligence and Research Group:
Hunting for Hackers
• Focus -- Creating leading threat intelligence
• Aggregation and analysis of unrivaled telemetry data at Cisco,
encompassing:
• Billions of web requests and emails
• Millions of malware samples
• Open source data sets (snort, clamAV…)
• Millions of network intrusions
56

CiscoTalos use ELK to analyze…
• Sandbox data cluster
• Dynamic malware analysis reports
• Search for related pattern, malewares
• ES stats
• 10 nodes
• 3 TB
• 100k reports/day
• ~8 months of data
• Honeypot cluster
• Collect attackers’ attempt
• { Account, password } pair
• Executed commands
• url of download files
• Suspicious command center for report back
57

Yale’s {elastic}SEARCH –
The Search for Cancer’s Causes and Cures
https://www.elastic.co/elasticon/2015/sf/videos/search-for-cancer-causes-and-cures
• With Next generation
sequencing technology, the lab
can process 8 million patients
specimens yearly
• How to interpret this amount
of data  what software can be
used
58

NYC restaurants inspection @ELK
• Real data from NYC open data project
• Restaurants inspection data
• Restaurants info
• Inspection info
• Violation codes and description
60

ELK stack introduction

Recommended

More Related Content

What's hot (20)

Similar to ELK stack introduction (20)

More from abenyeung1 (9)

Recently uploaded (16)

ELK stack introduction