Andrey Vasnetsov

> after 100 vectors with 800 limit and without payload I got the same error. Do I get it right, that you are making a single batch request with 100...

> I did not catch "size of data" vector size and number of vetctors

are you building a dev branch?

If you are using mmap, it might be needed to heat-up disk cache. So the first queries are going to be slow. We will look into options of how to...

yes, it is about right: ``` 5000000 x 768 x 4 - original vectors + 5000000 x 768 - quantized + ~30% indexes ``` Qdrant always keeps the original vectors,...

ToDo: bechmark the cloud with advanced configuration - io_uring + hybrid disks

Hi @nilhex, sorry for delay, I will come back to this task on the next week hopefully. Thanks for the contribution anyway!

hi @nairajay2k , do you have an estimation of how much data you upload in one batch?

is it possible to split batch into smaller pieces? Otherwise you can increase timeout.

https://github.com/qdrant/qdrant-client/blob/master/qdrant_client/qdrant_client.py#L69