thread 'solana-record-stage' panicked at 'assertion failed: serialized_size(&entry).unwrap() <= BLOB_DATA_SIZE as u64', src/entry.rs:67:9

[mvines]

This assert is failing on the testnet leader sometimes:
https://github.com/solana-labs/solana/blob/73ae3c3301805/src/entry.rs#L67

thread 'solana-record-stage' panicked at 'assertion failed: serialized_size(&entry).unwrap() <= BLOB_DATA_SIZE as u64', src/entry.rs:67:9
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
stack backtrace:
   ...
   7: solana::entry::Entry::new
   8: solana::ledger::next_entries_mut
   9: solana::record_stage::RecordStage::process_signal

[mvines]

New info from #774:

thread 'solana-record-stage' panicked at 'Serialized entry size too large: 66204 (292 transactions):', src/entry.rs:70:13

[rob-solana]

lots of votes, I guess!

[mvines]

The transactions should be getting chunked well before 292 though! https://github.com/solana-labs/solana/blob/master/src/ledger.rs#L77-L79

[rob-solana]

yep, I don't see how 292 gets through unless that vector of transactions is being modified while being passed to recorder

[mvines]

My current guess is that chunk_len = 1 here [1].

[1] https://github.com/solana-labs/solana/blob/master/src/ledger.rs#L82

[rob-solana]

I don't get it. chunk_len = 1 would mean a lot of single-Transaction Entries, right?