High Availability File Adapter in OSB

While running the FileAdapter (eis/FileAdapter) in a cluster with 2 instances, I get this error

BINDING.JCA-11042 File deletion failed

the full error is

onFatalError: Adapter forced endpoint deactivation due to:
BINDING.JCA-11042
File deletion failed.

here https://forums.oracle.com/forums/thread.jspa?threadID=2157695 and here https://kr.forums.oracle.com/forums/thread.jspa?messageID=9757727 and here https://cn.forums.oracle.com/forums/thread.jspa?threadID=2252967 they report the same story.


OK, we need to use eis/HAFileAdapter and give a controlDir. But we are not having a SOAINFRA DB schema because we installed only OSB.

see High Availability in Inbound Operations here http://download.oracle.com/docs/cd/E15523_01/integration.1111/e10231/adptr_file.htm

it is not clear if you have an option to use a file as a coordinator, instead of a DB. What is controlDir for?

http://download.oracle.com/docs/cd/E14571_01/doc.1111/e17059.pdf here the HA FileAdapter documentation


This document is also quite cool, it says:

"Database-based mutex and locks are used to coordinate these operations in a File Adapter clustered topology. Other coordinators are available but Oracle recommends using the Oracle Database."

yet they don't mention the alternatives.... if you use a DB, then it must be a HA DB otherwise a DB failure would mean a File Adapter failure...



see also here for full solution

Oracle High Availability Assessment with Failure Injection

A very interesting document (aka "MAA SOA Assessment") on failover for these configurations:

Mediator and JMS Adapter
BPEL and Oracle Database Adapter
BPEL and File Adapter


http://www.oracle.com/technetwork/database/features/availability/maa-soa-assesment-194432.pdf


on all sort of wounds inflicted to a HA Oracle env:

WLS Server Failure,
WLS Server Migration,
Node failure,
DB Instance failure,
DB node failure,
File System failure

and the the resiliency of the system is assessed.

( see also my "Break and Seek" post)


Another very interesting reading on the migration and failure recovery in WebLogic is here