feat: add region_statistics table (#4771)

* refactor: introduce `region_statistic`

* refactor: move DatanodeStat related structs to common_meta

* chore: add comments

* feat: implement `list_region_stats` for `ClusterInfo` trait

* feat: add `region_statistics` table

* feat: add table_id and region_number fields

* chore: rename unused snafu

* chore: udpate sqlness results

* chore: avoid to print source in error msg

* chore: move `procedure_info` under `greptime` catalog

* chore: apply suggestions from CR

* Update src/common/meta/src/datanode.rs

Co-authored-by: jeremyhi <jiachun_feng@proton.me>

---------

Co-authored-by: jeremyhi <jiachun_feng@proton.me>

src/datanode/Cargo.toml

@@ -53,6 +53,7 @@ prost.workspace = true
 query.workspace = true
 reqwest.workspace = true
 serde.workspace = true
+serde_json.workspace = true
 servers.workspace = true
 session.workspace = true
 snafu.workspace = true

src/datanode/src/heartbeat.rs

@@ -12,11 +12,13 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+use std::collections::HashMap;
 use std::sync::atomic::{AtomicBool, Ordering};
 use std::sync::Arc;
 use std::time::Duration;
 
 use api::v1::meta::{HeartbeatRequest, NodeInfo, Peer, RegionRole, RegionStat};
+use common_meta::datanode::REGION_STATISTIC_KEY;
 use common_meta::distributed_time_constants::META_KEEP_ALIVE_INTERVAL_SECS;
 use common_meta::heartbeat::handler::parse_mailbox_message::ParseMailboxMessageHandler;
 use common_meta::heartbeat::handler::{
@@ -320,16 +322,25 @@ impl HeartbeatTask {
         region_server
             .reportable_regions()
             .into_iter()
-            .map(|stat| RegionStat {
-                region_id: stat.region_id.as_u64(),
-                engine: stat.engine,
-                role: RegionRole::from(stat.role).into(),
-                // TODO(weny): w/rcus
-                rcus: 0,
-                wcus: 0,
-                approximate_bytes: region_server.region_disk_usage(stat.region_id).unwrap_or(0),
-                // TODO(weny): add extensions
-                extensions: Default::default(),
+            .map(|stat| {
+                let region_stat = region_server
+                    .region_statistic(stat.region_id)
+                    .unwrap_or_default();
+                let mut extensions = HashMap::new();
+                if let Some(serialized) = region_stat.serialize_to_vec() {
+                    extensions.insert(REGION_STATISTIC_KEY.to_string(), serialized);
+                }
+
+                RegionStat {
+                    region_id: stat.region_id.as_u64(),
+                    engine: stat.engine,
+                    role: RegionRole::from(stat.role).into(),
+                    // TODO(weny): w/rcus
+                    rcus: 0,
+                    wcus: 0,
+                    approximate_bytes: region_stat.estimated_disk_size() as i64,
+                    extensions,
+                }
             })
             .collect()
     }

src/datanode/src/region_server.rs

@@ -54,7 +54,7 @@ use snafu::{ensure, OptionExt, ResultExt};
 use store_api::metric_engine_consts::{
     FILE_ENGINE_NAME, LOGICAL_TABLE_METADATA_KEY, METRIC_ENGINE_NAME,
 };
-use store_api::region_engine::{RegionEngineRef, RegionRole, SetReadonlyResponse};
+use store_api::region_engine::{RegionEngineRef, RegionRole, RegionStatistic, SetReadonlyResponse};
 use store_api::region_request::{
     AffectedRows, RegionCloseRequest, RegionOpenRequest, RegionRequest,
 };
@@ -312,9 +312,9 @@ impl RegionServer {
         self.inner.runtime.clone()
     }
 
-    pub fn region_disk_usage(&self, region_id: RegionId) -> Option<i64> {
+    pub fn region_statistic(&self, region_id: RegionId) -> Option<RegionStatistic> {
         match self.inner.region_map.get(&region_id) {
-            Some(e) => e.region_disk_usage(region_id),
+            Some(e) => e.region_statistic(region_id),
             None => None,
         }
     }

src/datanode/src/tests.rs

@@ -31,7 +31,9 @@ use query::query_engine::{DescribeResult, QueryEngineState};
 use query::{QueryEngine, QueryEngineContext};
 use session::context::QueryContextRef;
 use store_api::metadata::RegionMetadataRef;
-use store_api::region_engine::{RegionEngine, RegionRole, RegionScannerRef, SetReadonlyResponse};
+use store_api::region_engine::{
+    RegionEngine, RegionRole, RegionScannerRef, RegionStatistic, SetReadonlyResponse,
+};
 use store_api::region_request::{AffectedRows, RegionRequest};
 use store_api::storage::{RegionId, ScanRequest};
 use table::TableRef;
@@ -210,7 +212,7 @@ impl RegionEngine for MockRegionEngine {
         unimplemented!()
     }
 
-    fn region_disk_usage(&self, _region_id: RegionId) -> Option<i64> {
+    fn region_statistic(&self, _region_id: RegionId) -> Option<RegionStatistic> {
         unimplemented!()
     }