records number of residual data shreds which don't make a full batch (#…

…25693)

Data shreds are batched into MAX_DATA_SHREDS_PER_FEC_BLOCK shreds for
each erasure batch. If there are residual shreds not making a full
batch, then we cannot generate coding shreds and need to buffer shreds
until there is a full batch; This may add latency to coding shreds
generation and broadcast.
In order to evaluate upcoming changes removing this buffering logic,
this commit adds metrics tracking residual number of data shreds which
don't make a full batch.

ledger/src/shred/stats.rs

@@ -1,4 +1,5 @@
 use {
+    crate::shred::MAX_DATA_SHREDS_PER_FEC_BLOCK,
     solana_sdk::clock::Slot,
     std::{
         ops::AddAssign,
@@ -17,6 +18,9 @@ pub struct ProcessShredsStats {
     pub sign_coding_elapsed: u64,
     pub coding_send_elapsed: u64,
     pub get_leader_schedule_elapsed: u64,
+    // Number of data shreds from serializing ledger entries which do not make
+    // a full batch of MAX_DATA_SHREDS_PER_FEC_BLOCK; counted in 4 buckets.
+    num_residual_data_shreds: [usize; 4],
 }
 
 #[derive(Default, Debug, Eq, PartialEq)]
@@ -43,6 +47,9 @@ impl ProcessShredsStats {
         let slot_broadcast_time = slot_broadcast_time
             .map(|t| t.as_micros() as i64)
             .unwrap_or(-1);
+        self.num_residual_data_shreds[1] += self.num_residual_data_shreds[0];
+        self.num_residual_data_shreds[2] += self.num_residual_data_shreds[1];
+        self.num_residual_data_shreds[3] += self.num_residual_data_shreds[2];
         datapoint_info!(
             name,
             ("slot", slot, i64),
@@ -60,9 +67,35 @@ impl ProcessShredsStats {
             ("gen_coding_time", self.gen_coding_elapsed, i64),
             ("sign_coding_time", self.sign_coding_elapsed, i64),
             ("coding_send_time", self.coding_send_elapsed, i64),
+            (
+                "residual_data_shreds_08",
+                self.num_residual_data_shreds[0],
+                i64
+            ),
+            (
+                "residual_data_shreds_16",
+                self.num_residual_data_shreds[1],
+                i64
+            ),
+            (
+                "residual_data_shreds_24",
+                self.num_residual_data_shreds[2],
+                i64
+            ),
+            (
+                "residual_data_shreds_32",
+                self.num_residual_data_shreds[3],
+                i64
+            ),
         );
         *self = Self::default();
     }
+
+    pub(crate) fn record_num_residual_data_shreds(&mut self, num_data_shreds: usize) {
+        const SIZE_OF_RESIDUAL_BUCKETS: usize = (MAX_DATA_SHREDS_PER_FEC_BLOCK as usize + 3) / 4;
+        let residual = num_data_shreds % (MAX_DATA_SHREDS_PER_FEC_BLOCK as usize);
+        self.num_residual_data_shreds[residual / SIZE_OF_RESIDUAL_BUCKETS] += 1;
+    }
 }
 
 impl ShredFetchStats {
@@ -99,6 +132,7 @@ impl AddAssign<ProcessShredsStats> for ProcessShredsStats {
             sign_coding_elapsed,
             coding_send_elapsed,
             get_leader_schedule_elapsed,
+            num_residual_data_shreds,
         } = rhs;
         self.shredding_elapsed += shredding_elapsed;
         self.receive_elapsed += receive_elapsed;
@@ -108,5 +142,8 @@ impl AddAssign<ProcessShredsStats> for ProcessShredsStats {
         self.sign_coding_elapsed += sign_coding_elapsed;
         self.coding_send_elapsed += coding_send_elapsed;
         self.get_leader_schedule_elapsed += get_leader_schedule_elapsed;
+        for (i, bucket) in self.num_residual_data_shreds.iter_mut().enumerate() {
+            *bucket += num_residual_data_shreds[i];
+        }
     }
 }

ledger/src/shredder.rs

@@ -153,6 +153,7 @@ impl Shredder {
 
         process_stats.serialize_elapsed += serialize_time.as_us();
         process_stats.gen_data_elapsed += gen_data_time.as_us();
+        process_stats.record_num_residual_data_shreds(data_shreds.len());
 
         data_shreds
     }