What is a stable Couchbase setup for 4.0 beta?

How do you suggest that we set up to test and run CB 4.0 with samples ?

A)
1 node with all on it Data, Query , index

B)
2 nodes with all on it Data, Query , index

C)
e.g.
1 node for Data
1 node Query
1 node index

D)
1 node with data and query
1 node with index and query
1 node with index and query

I ask this as I think that the clustering is not working correctly in CB 4.0 beta I have already posted issue

I need a basic set up to test any help is welcome ?

Hi @GeorgeLeon,
Sorry for the delay I will find an answer for you.

Regarding what option is best, mostly depends on what scenario you would like to test.
If you would like to test/tune for query performance with N1QL, then this might help:
http://www.couchbase.com/nosql-resources/presentations/tuning-query-performance-with-n1ql-in-couchbase-server-4.0.html

For a more general introduction to the options to scale the node services separately then this “hangout” video with Chian is a very good intro:

I hope this helps

Thanks @martinesmann

It really helps it is very good video by Cihan he is great at explaining and your all doing very good job with Couchbase
as it is beginning of new world order in database world I understand the challenges of cutting edge technology
I could only get so far by my own and i ended having to ask for some help I hope is ok.
Any feed back from anyone who can understand the errors would help to point us in the right direction .

Anyway be good and if anyone from CB comes to Greece look me up.

@GeorgeLeon
Right now the recommended setup is a single node cluster with everything enabled. In your case that would mean option A.

To get the most out of the single node cluster I would recommend that you configure the node above the minimum recommended size, witch is 4 cores, 8GB Ram.
Personally I run tests on a 8 core + 14 GB ram ( Azure VM A4, http://azure.microsoft.com/en-us/pricing/details/virtual-machines/)

This will of course change when the RTM version is released.

And you are always welcome to ask for help! :slight_smile:

@martinesmann

thanks will do