I need low latency search feature to query for say…texts within documents from 8 years ago. I also need sharding, load balancing, cluster awareness etc. Im looking for something open source as I was searching for alternatives for elasticsearch. Is this a better solution, if I could get an honest opinion?
Im currently running clusters of nosql (cassandra/scylladb) most of which has a schema.
Also, whats the situation of integrating AI into manticore for knowledge generation, analysis, & searching features like chatgpt?
Also, if possible id like a benchmark link comparing it with tantivy & typesense
Manual available. Autosharding - in progress (see roadmap.manticoresearch.com)
Yes. Manticore Search Manual: Creating a cluster > Remote nodes > Load balancing
Not sure what exactly you mean. Routing based on a value in the document - no.
situation of integrating AI into manticore
Vector search support is in progress HNSW Support for Vector Search · Issue #1415 · manticoresoftware/manticoresearch · GitHub
searching features like chatgpt
No such plans. Just regular vector search for now.
benchmark link comparing it with tantivy & typesense
Not available yet as far as I know.
thankyou for the help.
Really appreciate it. Ill look into it
Also, just 1 last thing if you dont mind-
If I need indexing for large datasets, then elasticsearch typically likes bulk insertions using the bulk API. Does that stand true for manticore too? Bulk insertions in chunks or can I do real time indexing? Whats more resource efficient considering a load of say 100k writes/sec/cluster
& for a language with no clients, should I just connect to it via 9312 & send binaries using manticore binary API? since im using a nosql db primarily with distributed nodes/clusters