diff --git a/ledger/src/blockstore.rs b/ledger/src/blockstore.rs
index ff14c9d5500d66..c5dcd7236e0707 100644
--- a/ledger/src/blockstore.rs
+++ b/ledger/src/blockstore.rs
@@ -1420,6 +1420,8 @@ impl Blockstore {
         // We don't want only a subset of these changes going through.
         write_batch.put_bytes::<cf::ShredData>(
             (slot, index),
+            // Payload will be padded out to SHRED_PAYLOAD_SIZE
+            // But only need to store the bytes within data_header.size
             &shred.payload[..shred.data_header.size as usize],
         )?;
         data_index.set_present(index, true);
@@ -1452,6 +1454,9 @@ impl Blockstore {
         use crate::shred::SHRED_PAYLOAD_SIZE;
         self.data_shred_cf.get_bytes((slot, index)).map(|data| {
             data.map(|mut d| {
+                // Only data_header.size bytes stored in the blockstore so
+                // pad the payload out to SHRED_PAYLOAD_SIZE so that the
+                // erasure recovery works properly.
                 d.resize(cmp::max(d.len(), SHRED_PAYLOAD_SIZE), 0);
                 d
             })
diff --git a/ledger/src/shred.rs b/ledger/src/shred.rs
index 3f8f41b39c0329..695105d4215aa1 100644
--- a/ledger/src/shred.rs
+++ b/ledger/src/shred.rs
@@ -319,11 +319,10 @@ impl Shred {
             Self::deserialize_obj(&mut start, SIZE_OF_COMMON_SHRED_HEADER, &payload)?;
 
         let slot = common_header.slot;
-        let expected_data_size = SHRED_PAYLOAD_SIZE;
         // Shreds should be padded out to SHRED_PAYLOAD_SIZE
         // so that erasure generation/recovery works correctly
         // But only the data_header.size is stored in blockstore.
-        payload.resize(expected_data_size, 0);
+        payload.resize(SHRED_PAYLOAD_SIZE, 0);
         let shred = if common_header.shred_type == ShredType(CODING_SHRED) {
             let coding_header: CodingShredHeader =
                 Self::deserialize_obj(&mut start, SIZE_OF_CODING_SHRED_HEADER, &payload)?;