256MB is min required for GSI indexes on ForestDB storage. More memory is better though. If you are using Memory optimized Indexes (MOI) then whole index need to fit in memory.
Can you post the errors you are hitting, and details such as couchbase version number, size of documents, indexes etc (the typical questions you see in the sizing guide).
I created a new index and query service CPU spiked up to 99%. The entire couchbase slowed down for the next 15 minutes.
At this point, I’m thinking about dropping all n1ql queries and go back to views since it is guaranteed to have at least one issue whenever I create a new index. We had the following issues with n1ql index so far.
Index hangs at x%. Only way to fix this was to drop it.
Query service CPU spike up to 99% and stays for while
index service restarts.
Can I get size guideline for n1ql PLEASE? If you need more information than the following, I’m more than happy to share.
Item Count: 47,250,214
Document size varies.
Currently assigned ram to the indexer: 12288 MB
number of index servers: 2
number of query servers: 2
@moon0326. I run WW professional services for couchbase.
Would love to spend some time understanding your requirements and your configuration so that we can provide appropriate guidance.