Loki Questions about bloom configuration

[李云中]

I saw on the official website that Bloom configuration does not require specifying a key or label. By default, all logs will be split according to Loki’s own rules and Bloom configuration will be done for all logs. For example, for JSON logs, Loki will parse the key and value and perform Bloom configuration. Will entering the key or value when I query have Bloom acceleration effect for all logs? If this really feels unreasonable, I think the best way is to design indexes for all labels, and Bloom can be considered as an index