GO BEYOND DATA
Real-time Analytics for Application Performance Management
Yury Oleynik Data Analyst
GO BEYOND DATA Real-time Analytics for Application Performance - - PowerPoint PPT Presentation
GO BEYOND DATA Real-time Analytics for Application Performance Management Yury Oleynik Data Analyst Modern applications Agenda Monitoring challenges INSTANA apploach 2 Instana, Inc. Proprietary and Confidential Business Demand: Write and
Yury Oleynik Data Analyst
Instana, Inc. Proprietary and Confidential
2
Modern applications Monitoring challenges INSTANA apploach
Instana, Inc. Proprietary and Confidential
3
Business Demand: Write and Deploy code faster!
Drives adoption of
Drives adoption of
Instana, Inc. Proprietary and Confidential
4
Monitoring
Why monitor systems and the applications?
Reality of monitoring
How it should be
January 2014
“And at our scale, humans cannot continuously monitor the status of all of our systems. To maintain high availability across such a complicated system, and to help us continuously improve the experience for our customers, it is critical for us to have exceptional tools coupled with intelligent analysis to proactively detect and communicate system faults and identify areas of improvement.”
Instana, Inc. Proprietary and Confidential
6
Daily/Hourly code and configuration changes.
Monitoring Challenges: Intense operational complexity
Shared infrastructure.
Throw away infrastructure.
Experts knowledge needed.
Non-deterministic code path.
Instana, Inc. Proprietary and Confidential
7
Instana, Inc. Proprietary and Confidential
8
Creating a Virtual DevOps Expert
Instana, Inc. Proprietary and Confidential
9
Creating a Virtual DevOps Expert
Intelligent Sensor Technology
Instana, Inc. Proprietary and Confidential
Data viewed as 5 second running average followed by 1 second data points. INSTANA collects 1 second resolution data. Data viewed as 1 minute running average
Aggregation = loss of information
10
CURRENT APM
Intelligent Sensor Technology Realtime Sensoring
Instana, Inc. Proprietary and Confidential
11
Creating a Virtual DevOps Expert
Intelligent Health Management
Instana, Inc. Proprietary and Confidential
12 Persistence Stream Processing Dynamic Graph Communication Sensor Data
Realtime Stream Processing
Machine Learning Knowledge Base Predictive Alerting
Health Management
Raw Store
Memory
Result Store Health Signatures 3D Map
Intelligent Health Mangement Streaming, Analytics, Learning & Knowledge
Instana, Inc. Proprietary and Confidential
13
Dynamic Dependency Graph
JV JV JV JV JV JV JV JV JVCassandra Cluster
Service
A
Tomcat Cluster
JVM JVM
App
1
App
2 JVM JVM
App
3
Service
B
Schema1
Schema2
Instana, Inc. Proprietary and Confidential
JV JV JV JV JV JV JV JV JVCassandra Cluster
Service
A
Tomcat Cluster
JVM JVM
App
1
App
2 JVM JVM
App
3
Service
B
Schema1
Schema2
JVCassandra Cluster Cassandra Cluster
Dynamic Dependency Graph
14
Instana, Inc. Proprietary and Confidential
JV JV JV JV JV JV JV JV JVCassandra Cluster
Service
A
Tomcat Cluster
JVM JVM
App
1
App
2 JVM JVM
App
3
Service
B
Schema1
Schema2
Cassandra Cluster Cassandra Cluster
15
Dynamic Dependency Graph
App
2
JVInstana, Inc. Proprietary and Confidential
16
Predictive Alerting & Optimization
Realtime Event Stream Intelligent Health Management Alterts & Optimizations
Severe Situation Detected
JVM GC Overhead too high - Impact on Service >20%. Knowledge Base recommendations:
at 10:43am.
Optimization Detected.
com.mycomp.Calc.calc() consumes 20% of Clock time. Optimization will have high impact on Service response time for Shop service. a few seconds ago
Instana, Inc. Proprietary and Confidential
17
Creating a Virtual DevOps Expert
3D Map
Demo
Instana, Inc. Proprietary and Confidential
19
STAN - a Virtual DevOps Expert
Instana, Inc. Proprietary and Confidential
20