You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Along came Beats to the ELK family, giving us the ELK-B(ee)
Elastic Universe circa 2018
Licensing
Elastic is distributed under an Apache 2 “open core” model
X-Pack features require a license purchase
The source code for X-Pack is published, but under a restrictive license
Elastic has an interesting philosophy on openness
Elasticsearch
First and foremost, a search engine powered by Apache Lucene
Primarily an append-only database
Schema is good, but searching all the things covers for a lot of mistakes
Elastic rigorously pursues reliability and scalability
Scales through sharding and replication
Shards - split incoming data over multiple files, potentially on several nodes
Replication - copy data files over multiple nodes
Sharding improves write performance
Replication improves read performance
Primarily accessed via a REST API
Somebody realized early on that feeding log files into a common database and searching them was handy
Observability
APM - embed an agent into your web app to report transaction performance details
Logs - collect your application logs into one central database for searching and reporting
Metrics - collect numeric data about host performance including CPU, memory, disk, and network
Full Text
Site search - crawl your web site, and build a Google-like index
Workspace search - connect your cloud apps for unified search across Google Docs, Salesforce, ServiceNow, Sharepoint, JIRA, etc
Security
Endpoint - anti-virus agent runs on machines, gathers data, and applies policy
SIEM - aggregate security events from endpoint machines and network flow data
Not mentioned on the homepage - geospatial
Store points and polygons
Search by distance or intersection
Logstash
Intermediary between input data and Elasticsearch database
ETL (extract, transform, load) pipeline for data cleanup
Written in JRuby, but being converted to full Java
Skipping it tonight
Logstash Inputs
azure_event_hubs
generator
jms
redis
syslog
beats
github
jmx
relp
tcp
cloudwatch
google_pubsub
kafka
rss
twitter
couchdb_changes
graphite
kinesis
s3
udp
dead_letter_queue
heartbeat
log4j
salesforce
unix
elasticsearch
http
lumberjack
snmptrap
varnishlog
exec
http_poller
meetup
sqlite
websocket
file
imap
pipe
sqs
wmi
More Logstash inputs
ganglia
irc
puppet_facter
stdin
xmpp
gelf
jdbc
rabbitmq
stomp
Logstash Codecs - decode incoming data
avro
fluent
netflow
cef
graphite
nmap
cloudfront
gzip_lines
plain
collectd
json
protobuf
dots
json_lines
rubydebug
edn
line
edn_lines
msgpack
es_bulk
multiline
Logstash filters - transform data
aggregate
dissect
geoip
metricize
throttle
alter
dns
grok
metrics
tld
cidr
drop
i18n
mutate
translate
cipher
elapsed
jdbc_static
prune
truncate
clone
elasticsearch
jdbc_streaming
range
urldecode
csv
environment
json
ruby
useragent
date
extractnumbers
json_encode
sleep
uuid
de_dot
fingerprint
kv
split
xml
Beats
Lightweight data shipping agents
Written in GoLang
Static binary all the things!
Easy cross-compilation
Uniform configuration due to libbeat at the core
Beats
Packetbeat
apexbeat
collectbeat
fastcombeat
etcdbeat
Filebeat
burrowbeat
connbeat
flowbeat
execbeat
Winlogbeat
hsnburrowbeat
consulbeat
gabeat
factbeat
Metricbeat
cloudflarebeat
dockbeat
gcsbeat
fastcombeat
Heartbeat
cloudfrontbeat
Elasticbeat
githubbeat
flowbeat
Auditbeat
cloudtrailbeat
etcdbeat
gpfsbeat
gabeat
amazonbeat
cloudwatchmetricbeat
execbeat
hackerbeat
gcsbeat
apachebeat
cloudwatchlogsbeat
factbeat
Hsbeat
githubbeat
Beats
gpfsbeat
journalbeat
mqttbeat
packagebeat
redisbeat
hackerbeat
kafkabeat
mysqlbeat
phpfpmbeat
retsbeat
hsbeat
kafkabeat2
nagioscheckbeat
pingbeat
rsbeat
httpbeat
krakenbeat
nginxbeat
prombeat
saltbeat
hwsensorsbeat
lmsensorsbeat
nginxupstreambeat
prometheusbeat
serialbeat
icingabeat
logstashbeat
nsqbeat
protologbeat
springbeat
iobeat
mcqbeat
nvidiagpubeat
pubsubbeat
tracebeat
jmxproxybeat
mongobeat
openconfigbeat
redditbeat
twitterbeat
Beats
udpbeat
udplogbeat
unifiedbeat
uwsgibeat
varnishlogbeat
varnishstatbeat
vaultbeat
wmibeat
Kibana
Web front end to query and visualize Elasticsearch data
Written in Node.js
Three key things to note:
App picker
Time picker
Index picker
Sysmon - a brief aside
“System Monitor (Sysmon) is a Windows system service and device driver that, once installed on a system, remains resident across system reboots to monitor and log system activity to the Windows event log. It provides detailed information about process creations, network connections, and changes to file creation time. By collecting the events it generates using Windows Event Collection or SIEM agents and subsequently analyzing them, you can identify malicious or anomalous activity and understand how intruders and malware operate on your network.”