weblogic clustering

1) Database level failover clustering.

2) Application level load balancing clustering.

3) admin server will not be part of the cluster.

4) Cluster can not span domains.

5) Domain and below are clusters.

6) http clustering DNS round robin, Proxy server, Hard ware

7) DNS will not work if port number are different.

8) Production server does not use DNS round robin.

9) Proxy server on client is called proxy.

10) Proxy on server is called reverse proxy.

11) Proxy's problem is round robin algorithm.

12) People mostly use Hardware servers.

13) Big IP, IBM edge load balancer.

14) Replication groups for http sessions.

15) Replication group can have the limit of 2.

16) Clustering does not have any limits.

17) Persistent session-J2ee concept.

18) web.xml specify storage-type persistent and jdbc parameter.

19) Heartbeats called mulitcasting.

20) TCP have following types communication. Unicasting 1-1, Broad casting 1-all, Mulitcasting 1- many.

21) One adapter for internal and one adapter for external. Use access and mutlicast address.

22) Class D is dedicated to multicasting.....

23) end with 255

24) Class A, B, C is for unicasting.

1-126 class A

127 look back

128 -191 class B

192 - 223 class C

224 - 239 class D

25) Heart beats and JNDI updates (mulitcasting are required)

26) create server ---> Create Proxy ---> Assign server to cluster

27) Start Node manager.

28) Start clusters.

29) C:\devtools\bea10mp1user_projects\domains\cluster_demo_domain\bin>startManagedWebLogic.cmd Mem2 http://localhost:7001

30) PersistantStoreType parameter weblogic.xml param value replicated. (Replica aware).

31) Change it to jdbc.

32) Use both inmemory + persistent store.

33) JMS Server in a singleton. JMS Server -- target weblogic instance 1. Create them as migratable target. JDBC

-Order of loading classes.

1. Kernel

2. JDBC

3. JMS (System resources.)

4. jars

5. wars

6. startup classes.

7.applications.

Startupclasses