Couchbase no of index limitation

[hiren.rojasara]

Is there any index limitation on couchbase bucket/scope/collection level?

How many index should have in production server? or any client/case study you have?

Let say i want to create index like where custId = “c1”.

So here i want to create index per customerId, where there could be N no of customers. So can couchbase handle N no of indexes in same collection.

I have gone through below posts, but not helpful

Practical limits on the number of indexes

Hello, Here is my feedback about you question. Using GSI (Global secundary indexes) starting with 4.0, providme better perfomance than Views. For many reasons: They are global, not local to Vbuckets. This mean, that the key location, for range querys are much better, because the index has all the index entries ordered, thus a range scan is very faster. The index live in RAM memory, instead Views that live on disk structures. 4.5 provide more performance with MOI (Memory Optimized in Memory)…

N1QL Indexes Limit

Hello!! I’m implementing Couchbase Server 4.0 to manage the data storage of an instant messaging app I’m developing. I’ve arrived to the conclusion that the new indexing system of N1QL could help. In this particular case, I want to index user’s messages to make different queries on them (last read message, last synchronization date, etc.). Taking that into account, is it a good strategy to create an index for each conversation in my system? If so, performance would be affected by that? And, is …

[vsr1]

Index WHERE meant for unrelated document types.
If N is few you can create. If lot use as INDEX key leading most. If EE use partition BY HASH(cutsId) if it is immutable

Divide and Conquer: Couchbase GSI Index partitioning.

[mreiche]

You won’t have an index for new customers that are added. Also - I suspect there is no performance benefit. as the query will need to figure out which index to use - or use all of them. Why not use the customerId as the document id and use the kv api or “use keys”

[hiren.rojasara]

I have multiple records per customer in the same collection, so key is random guid.

If you can clarify below things:

1. is there hard coded limit for no of indexes? Yes / No ?
2. if there are many indexes, then if any document gets upserted, then will it take time indexes to be updated even if i have use partial index? Lets say i insert one record with custId = c1, then it will update only one partial index (where custId = c1) or affect others as well.

[mreiche]

1. I don’t know. One could perform an experiment to create indexes in a loop until failure. If there is no limit on the number of indexes, but you encounter other issues, then then the number of indexes would not be relevant. To determine if your design is feasible for your purpose, conduct an experiment.

Edit: 6708 and counting.

2. I don’t know why they would. Again, a quick experiment would reveal the behavior.

[vsr1]

If i am right around 1000 collections and 10000 indexes on cluster