Content

Target customers: All users. Features released: In compute-storage separation scenarios, data analysis performance deteriorates when DLA reads data from a remote data store over the network. For example, when you use DLA to read data from OSS, this operation has a higher latency than data reading from a local disk. In addition, OSS limits the maximum bandwidth that can be used by a single user. This increases the latency of data analysis. In the serverless SQL engine of DLA, a local cache mechanism is introduced to cache hot data on a local disk. This mechanism reduces the latency and removes limits on I/O when you read data from a remote data store, achieving low latency and high throughput.