Thanks for trying subdoc feature !
The two major benefits of subdoc are:
- Reduce network bandwidth because you don’t have to send/receive full document across the network.
- Latency improvement for get/set for large documents. In this case, only a small portion of the document needs to traverse the network, which reduces the latency.
Here’s the doc which talks about subdoc:
There’s no change in terms of the max size of document we supported in CB.
In terms of disk I/O and memory, there may not be much improvement because under the hood, we will need to persist the full document into disk.
In terms of data model, I think this will depend on your use cases:
- How often will the documents get updated?
- Do you only need to read one day’s events or you need to read multiple day’s events?
- What’s your latency requirements for get/set?
Please give it a try with DP 4.5 together with SDK for subdoc.
It will be great if you can provide some feedback after you give it a try.