Re: LookupCache, ServiceDiscoveryListener and discard guarantees

Rick Moynihan wrote:
> Are any assurances made by LookupCache that calls to discard()
> will result in serviceRemoved() method invocations prior to any calls to
> serviceAdded()/serviceChanged() involving the discarded service?

Technically, it's impossible for the appropriate orderings to be guarenteed,
because your view of the "lost" service is in no way tied to the LUS's view of a
  "discovered" service.  What happens when there is a "race condition" such as
this, is that the service registration will be missing for some period of time.

I don't use the lookup cache in my Jini applications.  Instead, I've implemented
by own in which I always initiate a lookup, after a discard by dropping the
listeners to the affected Registrars, and then adding them back, and performing
a lookup.

This allows me to always have a "concrete" view of the state of the system,
driven by the view that I desire to have.  I also then periodically requery the
LUS with the same lookup parameters to keep from getting bit by odd timing or
interesting failure modes that are really not solvable in 100% of the cases.

Gregg Wonderly

--------------------------------------------------------------------------
Getting Started:     http://www.jini.org/wiki/Category:Getting_Started
Community Web Site:  http://jini.org
jini-users Archive:  http://archives.java.sun.com/archives/jini-users.html
Unsubscribing:       email "signoff JINI-USERS"  to [email protected]