Managing HP Serviceguard for Linux Ninth Edition, April 2009

Authorization File Problems..................................................................................286
Timeout Problems..................................................................................................287
Messages................................................................................................................287
Lock LUN Messages....................................................................................................287
A Designing Highly Available Cluster Applications ......................................................................289
Automating Application Operation ................................................................................289
Insulate Users from Outages ......................................................................................290
Define Application Startup and Shutdown ................................................................290
Controlling the Speed of Application Failover ................................................................291
Replicate Non-Data File Systems ...............................................................................291
Evaluate the Use of a Journaled Filesystem (JFS)........................................................291
Minimize Data Loss ....................................................................................................291
Minimize the Use and Amount of Memory-Based Data ......................................292
Keep Logs Small ....................................................................................................292
Eliminate Need for Local Data ..............................................................................292
Use Restartable Transactions ......................................................................................292
Use Checkpoints .........................................................................................................293
Balance Checkpoint Frequency with Performance ...............................................293
Design for Multiple Servers ........................................................................................293
Design for Replicated Data Sites ................................................................................294
Designing Applications to Run on Multiple Systems .....................................................294
Avoid Node Specific Information ...............................................................................294
Obtain Enough IP Addresses ................................................................................295
Allow Multiple Instances on Same System ...........................................................295
Avoid Using SPU IDs or MAC Addresses .................................................................295
Assign Unique Names to Applications ......................................................................296
Use DNS ................................................................................................................296
Use uname(2) With Care ............................................................................................297
Bind to a Fixed Port ....................................................................................................297
Bind to Relocatable IP Addresses ...............................................................................297
Call bind() before connect() ...................................................................................298
Give Each Application its Own Volume Group .........................................................298
Use Multiple Destinations for SNA Applications ......................................................298
Avoid File Locking ......................................................................................................299
Restoring Client Connections ..........................................................................................299
Handling Application Failures ........................................................................................300
Create Applications to be Failure Tolerant .................................................................300
Be Able to Monitor Applications ................................................................................301
Minimizing Planned Downtime ......................................................................................301
Reducing Time Needed for Application Upgrades and Patches ...............................301
Provide for Rolling Upgrades ...............................................................................302
Do Not Change the Data Layout Between Releases .............................................302
Table of Contents 13