NEW FEATURES / ENHANCEMENTS: feat: support new SQL query syntax. feat: add new meta-database for keeping all the metadata in the cluster. feat: support event-driving stream computing. feat: provide a whole new message queue for sub/pub time series data. Feat: remove the restriction for several query syntax. enh: provide the cloud-native architecture of TDengine. enh: employ Raft protocol to ensure the consistency in distributed environment. enh: utilize the B+tree to keep the meta-data in vnode. enh: add invert index for the fast search of the tag data. enh: improve the launch speed significantly. enh: support keep 1 billion time series data in a cluster. enh: provide a configurable cache to hold the latest data with various types. enh: develop a new data transporter to between different nodes in the cluster. enh: utilize the distributed hash table to locate the table to variable nodes in the cluster. enh: support roll-up SMA for super table to down-sample the raw time series data. enh: support time interval SMA to provide the approximate the time window aggregate result. doc: provide an new user manual, including SQL syntax usage and cluster maintenance.