Tonano: a real project or an attack against the TON blockchain?

Introduction

A new project Tonano inspired by Bitcoin BRC20 Ordinals has arrived on TON and has been welcomed by the TON Community as the first major third party project with a real world use case. Even Rogozov, head of TOP one of the main developers on TON, praised the initiative.

The TON blockchain achieved a world record performance on 31 October and is considered one of the most scalable blockchains. However, a few hours after minting started, the blockchain started experiencing problems due to some bad hardware validators and became virtually unusable even though it had technically remained online. This is why it was perhaps unwise to postpone slashing optimisation mechanisms for inactive and underperforming validators from Q3 to Q4.

The TON blockchain is back up and running again after more than two days of problems.

Tonano project

Now we come to the Tonano project, which although it has a real use case that can attract many users, is full of shadows. Let us go in order:

1) The tonano.io website was registered at the end of November 2023 and the registration was updated on 5 December 2023. In addition, there is only one copy on wayback machine dated 5 December that contains an empty website.

2) Their X account has no activity before 4 December and only has about 4.2k followers.

3) Their Telegram group has no activity before 5 December and only has about 7k followers (now 4.8k probably the rest were bots).

4) The project is not present in any part of the TON ecosystem: no official channel, website or catalog.

5) Why did such a mysterious project that appeared out of nowhere and had few users decide to carry out a massive minting operation producing over 2 million transactions in one hour with just 20k unique addresses?

6) According to their documentation, each event must be sent to the black hole wallet, which is a wallet that is currently blocked according to the proposal of the TON Foundation and the vote of the validators. A project that has the ambitious goal of minting 2.1 billion inscriptions, but does not know how to use the TON blockchain properly, is quite curious.

7) Usually on TON it takes weeks, if not months, to approve a standard, for example the TEP standards. Tonano would like TON-20 to become the inscription token protocol standard. However, there is no active discussion on github.

8) The Tonano team claims to have stopped the minting and that it will work with the TON Core team before resuming. Why is it that if all services on TON are suspended, including Tonano, there are still transactions arriving in the validators queue, but more importantly, continuous transactions to the black hole wallet?

9) Following several questions in Tonano chat, the managers of the project stated that their team consists of three people, including one developer. It seems curious that such a team would be able to take down one of the best blockchains in existence with a team of professionals, designed by one of the most ingenious minds in the IT industry at the moment.

TON Foundation

Now it is the turn of the TON Foundation and its members who do not seem to agree on what the source of the problem was:

1) According to the TON Foundation, there was a problem due to the poor hardware of the validators and asked those with poor CPU not to participate. However, in a distributed system, the limiting factor is bandwidth rather than computing power (CPU/GPU) or memory (RAM or mass storage). System load has never gone beyond 3 and this number is accessible to any CPU in a modern laptop, even less onerous for that of a server. That is why this motivation also seems implausible.

2) The TON Foundation also mentioned insufficient bandwidth among the possible causes. The number of packets per second in the network never exceeded 15,500 pps which is a number compatible with about 10-200 Mbps Ethernet network (frame size 84-1542 Byte). Typically, data centres, even with the cheapest hosting profiles, have sufficient bandwidth for any blockchain, especially with the load of 3 million transactions that was present on TON at the time.

3) According to the data available from 18/02/23, during the period from 06/03/23 to 11/04/23 the CPU load and network traffic reached values comparable to those found during the performance problem from 05/12/23 to 11/12/23. Furthermore, even the number of validators (285-305) and their hardware were comparable during the two periods. So this also seems to refute the official hypothesis of the TON Foundation once again.

4) One of the members of the TON Foundation said that the bottleneck was in mass storage performance. According to the specifications, a TON validator node needs one NVMe SSD with at least 25k IOPS. However, even this assumption does not seem realistic as these values are achievable by even a SATA3 SSDs. If by IOPS they meant the 4KQD1 benchmark there are many entry level NVMe SSDs capable of achieving them. In general, data centres, even with the cheapest hosting profiles, provide this.

5) Some users claim that it could have been a serious software bug given the several updates to the network validators software before the problem was solved. Analysing the change logs (commit822, commit823), this hypothesis also seems to be rejected as they are apparently optimisations rather than bugs. Otherwise, it should have been clearly written, why wouldn't the TON core team have told the truth?

6) The version of the validator software with which TON broke the TPS world record is not yet used in the mainnet. Could this explain why, with 1,000 times less transaction volume, the blockchain suffered a major failure? It does not seem plausible. In fact, there was a performance test last March where with only 12 validators, 16 shards, a similar software version and hardware characteristics as at the time of the problem, the TON blockchain managed to handle loads between 700-2000 TPS. In any case, the TON Foundation should clarify why the version of the software that broke the TPS world record has not yet reached the mainnet.

Conclusions

Given the above analysis, a main question arises: does the Tonano project really exist or is it a cover for something else?

The justifications given by Tonano and the TON Foundation do not seem to match reality. Some hypotheses not yet refuted by the evidence:

1) Stress test. The Tonano project was created ad hoc to perform a stress test on the mainnet in order to highlight any problems and be ready for larger and more onerous projects to come e.g. Notcoin. Thanks to this situation, the TON Foundation was able to take a census of the validators in the network and exclude the unreliable ones (down from 342 to 259).

2) Undisclosed software bug. The particular type of highly unbalanced load towards a single wallet generated by the Tonano project revealed a bug in the validators software that had not previously been detected during the various tests, including that of the world record performance on 31 October. The TON Foundation used the situation to fix the bug without acknowledging it directly.

3) DDoS attack. Given the history and scope of the TON project, a DDoS attack by the only government that wants to maintain the status quo in the world could be behind the problem.

Based on the known facts and the evidence gathered, it is not possible to determine which hypothesis is the most plausible. However, what is certain is the lack of transparency of the TON Foundation, not limited to this particular circumstance (read more here about A Community Proposal for Tokenomics Optimization and TON Believers Fund).