08709_landlust haslam deployment patterns for fusion middleware 11g v2

8/12/2019 08709_Landlust Haslam Deployment Patterns for Fusion Middleware 11g v2

1/54

Deployment patterns

for Fusion Middlewarea best practice session bySimon Haslam & Jacco H. Landlust

1

maandag 10 oktober 2011


2/54

`id jacco`

Oracle Since 2000, Linux since 2001

Independent Red Stack Administrator Oracle ACE since 2006

Infrastructure: Architecture, High Availability,Troubleshooting & Performance

http://oraclemva.wordpress.com

2

http://oraclemva.wordpress.com/http://oraclemva.wordpress.com/


3/54

`id simon`

Oracle since 1996 (UNIX since 1989)

Founded Veriton in 1996 Oracle ACE Director since 2009

Arhitecture, Design, Installation

http://simonhaslam.co.uk

3

http://simonhaslam.co.uk/http://simonhaslam.co.uk/


4/54

Disclaimer

Best practices of our (Limited) experience

We are not working for Oracle

We do not pretend this list is complete

We are not native American speakers

4



5/54

Agenda

Best Practices for Designing a Fusion

Middleware 11g Infrastructure

5



6/54

Normal Design Process

1. Some product is bought (or build in-house) thatruns on FMW 11g

2. Buy some kit

3. a DBA is sent to OU to get training

4. The week after training DBA has to build and

maintain a production platform with help from oneor more developers

5. Start running production (and into trouble... )

6



7/54

Gartner

8 - 12 % TCO in Design + Develop + Test

88 - 92 % TCO in Maintaining

7



8/54

Middleware is not

Trivial

Its (mission) critical Complicated: lots of moving parts

Get involved in projects from the start

8



9/54

Domain

Node Manager Node Manager

Machine A Machine BAdminServer

Cluster

Managed Server Managed Server

Managed Server

9



10/54

Real Agenda

1. Logical Design

2. Physical Implementation

3. Maintaining, the Full Life Cycle

10



11/54

Logical Design

11



12/54

Decide upon

Definitions

Before you design the infrastructure Brief your whole team (includingmanagement)

Clear naming convention

12



13/54

Document your

Train of Thou ht Tedious but usefull for future use

Dont wait till after the project /implementation

Mark facts as facts and opinions as opinion

Formulas and rules

13



14/54

We Love Pictures

Node 2Node 1

SOA ClusterSOA

CRS

SOA

CRS

SAN 2SAN 1

ACFS

WLS WLS

If its to complicated too draw on

a diagram, its most likely wrong

14



15/54

Segregation between

Environments

DTAP & SOD

Separation of servers & network

15



16/54

Design for Security

SOD

SSL whenever possible Also in DEV

Firewall

Connection Filters

16



17/54

Dont mix Layered

Products in OneDomain

oracle_common scripts

dependencies on upgrade / patching

17



18/54

Nodemanager:

Domain vs Node

Node specific: settings are domain wide

Domain specific configuration

Separation of OS users

SSL trust-stores

18



19/54

Separate

System Componentsfrom

ava Com onents

Automatic network separation

Different purpose

19



20/54

Never run Apps in

AdminServer

Mixture of management tooling &customer applications

Difficult troubleshooting

20



21/54

Cluster Everything

From the start (one node cluster) Find unexpected behaviour because ofcluster as soon as possible

Extending clusters is easy

21



22/54

Separate Directories

Use personal accounts for administrators

Separate application users frominfrastructure users

22



23/54

Physical Implementation

23



24/54

x86

Weve only seen x86/x64 customersrunning Linux

Buy the fasted kit you can get

24



25/54

Scale-up vs

Scale-out

Larger JVM give more GC issues Our customers do scale-out

SAME

25



26/54

Virtualization

Not for vmotion-like functionality Cloning (if you put the effort into it)

Partition hardware

26



27/54

Operating System

Dont adopt a new OS because of WLS

The best OS is the one youre an expert at

27



28/54

Shared Storage

NFS

OCFS2 GFS2

ACFS

rsync

28



29/54

Shared Storage:

take notice of

HTTP lockfiles Large logfiles

UCM tmpdirs

29



30/54

Use VIPs

whenever Possible

Faster failover upon node failure Easier to rebuild on new node

Logical separation of application andmachine

30



31/54

Get a LoadBalancer

Performance Stability

SSL offloading

31



32/54

oraInventory in

MW HOME

oraInventory only used by installers &oPatch

self-contained package

easier backups

32



33/54

Separate Binaries

from Confi uration

Replace binaries more easy Local disk for binaries

Backup configuration

33



34/54

Only use Java-based

Nodemana er

Easier to setup

Lots of bugs exist for script-basednodemanager

SSL is a pain on script based nodemanager

(if not impossible)

34



35/54

Separate

AS from MS

On different boxes

Easier upgrade path

No interference of management tooling

(if you have the licenses)

35



36/54

Setup AdminChannel

and ort

Separate admin traffic from application Dedicated listen thread

Has to run over SSL

36



37/54

JDBC

min-connections = max-connections Statement cache size < open_cursors

FAN (GridLink) when on RACSee session #20942 13:15 on Wednesday

37



38/54

Centralize

Confi uration

No configfiles that need to be changed inJAR/EAR files

No environment dependent ear-files.

Easier to maintain

38



39/54

Logging

Centralize logging Logrotation on size + time

Easier troubleshooting

39



40/54

Standardize

Troughout all environments

Easier troubleshooting (at 2 AM)

40



41/54

Automate

Standardizes installation

No differences between prod and test

41



42/54

Test Test Test

Find bugs before go-live

Benchmark

Stress test until break

Predictable performance

42



43/54

Maintainingthe full life cycle

43



44/54

Track Changes

Audit provider

Record changes with WLST

44



45/54

Record Incidents

adrci:

- Build-in fault diagnostibility infrastructure

- Health monitor reports

- Package incidents & problem information

45



46/54

learn WLST

mBeanbrowser

Automating tasks

46



47/54

Manage deployments

Learn Ant

CloningClient.jar

No environment specific ear-files

Plan.xml

47



48/54

Out of PlaceUpgrades

48



49/54

Grid Control

Stack monitoring SLA checks

Corrective action library

49


MOS


50/54

MOS

Confi urations

Not OCM per-se

Faster SR handling

50



51/54

Summary

51



52/54

first Think, then Act

52



53/54

Standardize

&Automate

53



54/54

Document & Test

08709_landlust haslam deployment patterns for fusion middleware 11g v2

Documents