Apica - Performance Does Matter: Five Key Elements to Consider in the Cloud

Preview:

Citation preview

Excellence in Cloud Performance

Performance does matter5 things to consider

sven.hammar@apicasystem.comWatch the video of this presentation

2

Sites Continue to Crash

Extremely Sluggish Response Times

Why?

Situation Today

3

June 19, 2010

How to handle the expected?

4

How to prepare for the unexpected?...

5

Rapid Load Ramp-Up

Snowball Effect

Load Fixed Capacity Scalability Application Transparancy

Server & Cloud

Peak Load Problem

6

Tips & Suggestions

#1 For peak and high load Small is Fast Have backup plan “minimalistic start/landing pages“

#2 Extensive use of Front End Cache systems Optimize the cache solution, consider Varnish Less traffic is less problem, no direct DB access

#3 Implement Scaling & Queuing System Redirect excess trafic with LoadBalanser Create Informative ”Wait” pages

Be prepared : Test the solution before launch

7

Why Run A Performance Load Test ?

Is the site stable?

When does it crash?

How can I make it faster?

Can my application scale?

8

Load

vs.

Capacity

9

vs.

TestEnvironment

ProductionEnvironment

10

Throughput

Nr ofusers

Point of collapse

Completefailure

Response time∞

Load Maximum– How many users can we

handle ?– What is a good result ?

Behaviour in the ”Danger Zone”–

Does the application become unstable above load maximum ?

Problem Analysis– Where are the bottlenecks– How to fix them ?

Load test 1 to 1

Nr ofusers

11

Do you have Performance Targets?

We shall never crash due to loadWe shall be compareble with the best

in class sites for ...Our peak time response time shall be

better then site www.YYY

Level : We shall handle100.000 page views per hour with :

Better then 4 sec average response time

95% of our users shall make a selection for purchase of a (ticket, service etc .. ) in less then 30 seconds

12

LoadTest findings

#1 Identify the Backend Calls Database calls don’t kill your application Lack of caching does!

#2 Check Static Content Delivery Optimize the cache solution, consider Varnish Consider using a CDN, if needed

#3 Web Infrastructure Load Balancer Server model Bandwidth Scaling & Failover

13

Know your numbers

14

By the numbers

The need for a baseline My startpage, Login, Book a flight

Response Time – Average 3.2 sec Typical Values – Median 2.5 sec Standard Deviation 2.8 sec SLA % 99.9 95 % is better then 11,8 sec

15

Browser Scenarios Analytics Response time

Browser

Analytics Trend Complex

Application SLA on

applications

Application

Basic Monitor Alerting Up – down Basic SLA

UP/Down

Inside monitoring

Vital Signs Drill Down Root cause Consolidation

with other system

Correlation

WebPerformance Monitoring for the Cloud

1 2 3 4

16

17

Tips & Suggestions

#1 Set Goals Uptime Performance

#2 Hate the average Work with the exceptions Remove the 10 Worst transactions

every month

#3 Fire drill Help identifying problems Correlation of data

18

To sum it up

A load test before release …

1

Know your numbers...

2

Plan for the unexpected...

3

Fire Drill - Be prepared

4

All systems have a weak spot – what is yours ?

Thank You!

Excellence in Cloud Performance

20

Further information

www.apicasystems.com– www.apicasystems.com/partners/rightscale.aspx

sven.hammar@apicasystem.com

craig.irwin@apicasystem.com