2nd Level Caching Expected Behavior

[Carlos Mendes]

Hi,

What's the expected behavior if we are using a 2nd Level Caching Provider (eg. Velocity Provider - AppFabric Caching) and the Cache goes down?

Will NHibernate continue to serve the requests ignoring the caching configuration?

 

[Ayende Rahien]

It blows up, most frequently. 

The major problem in trying to silently recover from this is that just accessing the cache to figure out that it is down is a very expensive operation.

It is better to error quickly than try to recover.

--
You received this message because you are subscribed to the Google Groups "NHibernate Contrib - Development Group" group.
To post to this group, send email to nhc...@googlegroups.com.
To unsubscribe from this group, send email to nhcdevs+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/nhcdevs?hl=en.

[Carlos Mendes]

Thank you Ayende.

I've tested it today with the NH Redis provider if the cache goes down (I stopped the cache server) the provider throws exceptions and the application stops.

I was expecting NHibernate to ignore the cache configuration and serve the requests from the Database.

   

[Ayende Rahien]

See my earlier post on why this is actually NOT desriable.

   

--

[Carlos Mendes]

Yes I got it.

I'm just wondering how to proceed since Microsoft's App Fabric High Availability mode only runs on Windows 2008 Enterprise which is kind of expensive.

I'll look into Redis (with Master - Slave nodes) and Memcached to understand if I can get high availability without paying through the nose.

Thanks again