I am doing a POC for using Hibernate OGM with MongoDB and CouchDB (which I understand is in experimental stage as of now).
I had a couple of queries around it:
When can I expect OGM support for CouchDB in mainstream? Is it planned for near future?
I observed that when doing POC on a same hardware and OS platform (16GB RAM, dual core , Win 7x64 OS), OGM performs to satisfactory levels with mongodb (100K records inserted in around 25 seconds), as compared to couchdb (only 1000 records insertion took around 90 seconds). Is there any workaround to get a better performance from OGM with couchDB, or does it need any improvements at the driver level for couchDB?
Hi,
our team is small and we don’t have the resources right now to work on the CouchDB dialect. We would be happy to review and integrate PR if somebody wants to actively work on it. The amount of feedback we receive is also critical to figure out where to focus our energy.
About your question on performance, I don’t have an answer. We would need to run more benchmarks to figure that out. My guess is that there are areas in the dialect that can be improved. Anyway, if you have suggestions or find something, feel free to send us feedback.
I would be surprised that the code in OGM would make such a difference between 2 dialects. The dialects don’t have that much code.
Anyway, it seems that you have a self contained benchmark so it would be useful to attach VisualVM to it and run some sampling/profiling on it to see if the issue really comes from the dialect.