Jump to content


Continuous Queries and other features


  • Please log in to reply
No replies to this topic

#1 Guest_Hoerst_*

Guest_Hoerst_*
  • Guests

Posted 01 November 2010 - 08:26 PM

I read Iqbal Khan's article on MSDN and wondered, if the "continuous query" mentioned therein is available in NCache. And if so, can they be made durable, so that a client that goes down and logs into the cluster later will get the missed updates?

Also, can you register interest in items only by naming them verbatim, or is there something like prefix match or match by regex or similar?

As I could not find these features in the NCache docs on the website, I'm assuming they do not exist. A couple of other features I'd like to know more about to be able to evaluate this product and compare it to some of the feature-rich Java caches would be

* How does the custom event propagation work in more detail?
* Authorization details (key-based? per map? etc.)
* How are split-brain scenarios handled? (guaranteed consistency or focus on availability)
* Is the NCache embedded in the application or are those standalone processes?
* Is it possible to restart individual nodes without forcing (double) re-distribution of keys
* Is a consistent hashing algorithm being used for distribution?
* Transaction support?
* Support for rolling upgrades (both of the NCache software and/or the "schema" of the stored objects?)

It would be great to get specific details to these questions. Thanks in advance,
Hoerst




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users