Adds logging of column family size and database size on startup and s… (#8336)

* Adds logging of column family size and database size on startup and shutdown

* Changes log level of column families size strings to debug. Adds TODO comment to use human_bytes crate for human-readable format of metrics. Adds print_db_metrics function to ZebraDb struct.

* Calls enumerate() on column_families to access index var

* Resolves cargo fmt checker results

* Resolves clippy lint

* Runs and fixes changes from fmt, clippy, check and test

* Removes prop.txt

* minor doc changes

---------

Co-authored-by: Elijah Hampton <elijahhampton@MBP-de-Elijah-2.lan>
Co-authored-by: Pili Guerra <mpguerra@users.noreply.github.com>
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>

zebra-state/src/service.rs

@@ -254,6 +254,10 @@ impl Drop for StateService {
             "dropping the state: dropped unused non-finalized state queue block",
         );
 
+        // Log database metrics before shutting down
+        info!("dropping the state: logging database metrics");
+        self.log_db_metrics();
+
         // Then drop self.read_service, which checks the block write task for panics,
         // and tries to shut down the database.
     }
@@ -451,6 +455,11 @@ impl StateService {
         (state, read_service, latest_chain_tip, chain_tip_change)
     }
 
+    /// Call read only state service to log rocksdb database metrics.
+    pub fn log_db_metrics(&self) {
+        self.read_service.db.print_db_metrics();
+    }
+
     /// Queue a checkpoint verified block for verification and storage in the finalized state.
     ///
     /// Returns a channel receiver that provides the result of the block commit.
@@ -853,6 +862,11 @@ impl ReadStateService {
     pub fn db(&self) -> &ZebraDb {
         &self.db
     }
+
+    /// Logs rocksdb metrics using the read only state service.
+    pub fn log_db_metrics(&self) {
+        self.db.print_db_metrics();
+    }
 }
 
 impl Service<Request> for StateService {

zebra-state/src/service/finalized_state/disk_db.rs

@@ -13,6 +13,7 @@
 use std::{
     collections::{BTreeMap, HashMap},
     fmt::Debug,
+    fmt::Write,
     ops::RangeBounds,
     path::Path,
     sync::Arc,
@@ -21,7 +22,7 @@ use std::{
 use itertools::Itertools;
 use rlimit::increase_nofile_limit;
 
-use rocksdb::ReadOptions;
+use rocksdb::{ColumnFamilyDescriptor, Options, ReadOptions};
 use semver::Version;
 use zebra_chain::{parameters::Network, primitives::byte_array::increment_big_endian};
 
@@ -512,6 +513,55 @@ impl DiskWriteBatch {
 }
 
 impl DiskDb {
+    /// Prints rocksdb metrics for each column family along with total database disk size, live data disk size and database memory size.
+    pub fn print_db_metrics(&self) {
+        let mut total_size_on_disk = 0;
+        let mut total_live_size_on_disk = 0;
+        let mut total_size_in_mem = 0;
+        let db: &Arc<DB> = &self.db;
+        let db_options = DiskDb::options();
+        let column_families = DiskDb::construct_column_families(&db_options, db.path(), &[]);
+        let mut column_families_log_string = String::from("");
+        write!(column_families_log_string, "Column families and sizes: ").unwrap();
+        for cf_descriptor in column_families.iter() {
+            let cf_name = &cf_descriptor.name();
+            let cf_handle = db
+                .cf_handle(cf_name)
+                .expect("Column family handle must exist");
+            let live_data_size = db
+                .property_int_value_cf(cf_handle, "rocksdb.estimate-live-data-size")
+                .unwrap_or(Some(0));
+            let total_sst_files_size = db
+                .property_int_value_cf(cf_handle, "rocksdb.total-sst-files-size")
+                .unwrap_or(Some(0));
+            let cf_disk_size = live_data_size.unwrap_or(0) + total_sst_files_size.unwrap_or(0);
+            total_size_on_disk += cf_disk_size;
+            total_live_size_on_disk += live_data_size.unwrap_or(0);
+            let mem_table_size = db
+                .property_int_value_cf(cf_handle, "rocksdb.size-all-mem-tables")
+                .unwrap_or(Some(0));
+            total_size_in_mem += mem_table_size.unwrap_or(0);
+
+            // TODO: Consider displaying the disk and memory sizes in a human-readable format - #8380.
+            write!(
+                column_families_log_string,
+                "{} (Disk: {} bytes, Memory: {} bytes)",
+                cf_name,
+                cf_disk_size,
+                mem_table_size.unwrap_or(0)
+            )
+            .unwrap();
+        }
+
+        debug!("{}", column_families_log_string);
+        info!("Total Database Disk Size: {} bytes", total_size_on_disk);
+        info!(
+            "Total Live Data Disk Size: {} bytes",
+            total_live_size_on_disk
+        );
+        info!("Total Database Memory Size: {} bytes", total_size_in_mem);
+    }
+
     /// Returns a forward iterator over the items in `cf` in `range`.
     ///
     /// Holding this iterator open might delay block commit transactions.
@@ -720,6 +770,32 @@ impl DiskDb {
     /// <https://github.com/facebook/rocksdb/wiki/RocksDB-FAQ#configuration-and-tuning>
     const MEMTABLE_RAM_CACHE_MEGABYTES: usize = 128;
 
+    /// Build a vector of current column families on the disk and optionally any new column families.
+    /// Returns an iterable collection of all column families.
+    fn construct_column_families(
+        db_options: &Options,
+        path: &Path,
+        column_families_in_code: &[String],
+    ) -> Vec<ColumnFamilyDescriptor> {
+        // When opening the database in read/write mode, all column families must be opened.
+        //
+        // To make Zebra forward-compatible with databases updated by later versions,
+        // we read any existing column families off the disk, then add any new column families
+        // from the current implementation.
+        //
+        // <https://github.com/facebook/rocksdb/wiki/Column-Families#reference>
+        let column_families_on_disk = DB::list_cf(db_options, path).unwrap_or_default();
+        let column_families = column_families_on_disk
+            .into_iter()
+            .chain(column_families_in_code.iter().cloned())
+            .unique()
+            .collect::<Vec<_>>();
+        column_families
+            .into_iter()
+            .map(|cf_name| ColumnFamilyDescriptor::new(cf_name, db_options.clone()))
+            .collect()
+    }
+
     /// Opens or creates the database at a path based on the kind, major version and network,
     /// with the supplied column families, preserving any existing column families,
     /// and returns a shared low-level database wrapper.

zebra-state/src/service/finalized_state/zebra_db.rs

@@ -322,6 +322,14 @@ impl ZebraDb {
 
         Ok(())
     }
+
+    /// Logs metrics related to the underlying RocksDB instance.
+    ///
+    /// This function prints various metrics and statistics about the RocksDB database,
+    /// such as disk usage, memory usage, and other performance-related metrics.
+    pub fn print_db_metrics(&self) {
+        self.db.print_db_metrics();
+    }
 }
 
 impl Drop for ZebraDb {

zebrad/src/commands/start.rs

@@ -131,6 +131,9 @@ impl StartCmd {
             )
             .await?;
 
+        info!("logging database metrics on startup");
+        read_only_state_service.log_db_metrics();
+
         let state = ServiceBuilder::new()
             .buffer(Self::state_buffer_bound())
             .service(state_service);