It is Stressful Being a Cryptocurrency! (Byteball Stress Tests in 2018)
Blockchains are regularly put through rigorous testing by their founders and development teams to ensure their ability to handle scaling, malicious attacks, and the typical functionalities of the blockchain. However, because most cryptocurrencies are open source, developers around the world sometimes take the liberty to perform these tests as well. Byteball is a cryptocurrency that is undergoing many of these intentional stress tests in February 2018.
What is Byteball Bytes (GBYTE)?
Byteball Bytes (GBYTE) is a cryptocurrency with a market cap of approximately $405 million and an individual coin value of $628. GBYTE is predominantly traded on OKEx and Binance with fairly substantial volume. However, what makes GBYTE unique is their technology and their blockchain platform/concept. GBYTE has created a risk-free conditional smart payments system. Being able to place a specific condition on a payment and ensuring it is met before the currency is released is specifically what GBYTE specializes in. If the condition is not met in the specified period of time, the funds are returned to the sender.
The cryptocurrency world has been known for their inability to retrieve stolen funds, the mystique, and allure associated with privacy coins, or that even the most transparent coins are just a chain of letters and numbers that rarely can be linked to a person. GBYTE implements another conditional requirement to ensure funds are paid when conditions are met, or they will be rightfully returned to the sender.
Byteball’s Technology and Business Plan
The concept is fairly simple for GBYTE. The sender wishes to send money with specific conditions associated with it. This is actually a unique way to place a hedge against certain risks in the market. The conditions can be almost anything, “If bitcoin falls under $5,000 pay individual X, $20,000.” Specifically, the equivalent of insurance policies to hedge against negative events can be bought and sold. This entire business model is based on smart contract technology which is one of the driving forces behind GBYTE.
The conditional payment system for something like insurance is what makes GBYTE very unique. However, this system can also be used for predictive purposes. GBYTE users can theoretically bet against each other regarding future events under an enforceable contract system. An example of this would be two work colleagues, Individual A believing BTC will go to $20,000, while individual B believes BTC will go to $1,000. With BTC trading at $10,000 that is $9,000 down and $10,000 up. Individual A goes, “I’ll bet you $50 it hits $20,000 before $1,000.” Individual B accepts the offer, and both buy GBYTE to put the smart contract on the blockchain. Whichever event occurs first will automatically transfer the GBYTE that in essence has been held in escrow by smart contracts until the triggering event.
The real question to ask is not will it work because it already does. The question that remains to be asked is how the GBYTE platform will handle scaling issues as more and more smart contracts are generated? To answer this GBYTE was put through three major stress tests in February 2018.
It is Stressful Being a Cryptocurrency
Byteball’s blockchain has been tested three times in February. The first test occurred February 8 at 10 am UTC. The purpose was specifically to determine how the Byteball network would respond under high traffic conditions.
To understand the stress test GBYTE went through it is important to understand how the test was set up. Anton Churyumov is the lead dev and one of the founders of Byteball. He claims the average network transactions per second (tps) was approximately ten tps. However, this was not able to be verified, and a network stress test was invented. A bot was created specifically with the intention of creating maximum stress on the GBYTE network. The bot would stress the network by sending a large number of posts or a continuous flow of posts at fixed theoretical tps levels. While the bot was flooding the GBYTE network, it would simultaneously be logging the results for later analysis. GBYTE was aware of these stress tests, and prior to all tests, the bot acquired GBYTEs to be used during the stress tests.
Test 1: February 8, 2018
The first test of three tested the Byteball’s network response under conditions of being flooded with transactions. The stress bot successfully sent significant batches of posts to be published. The first was five, twenty post batches. This minimal amount was to get an average value for the remaining test. The second-round the stress bot sent three, fifty post batches, once again to get an average value. The experiment originally planned to increase batches, fifty posts at a time until two hundred and fifty was reached. The experiment was halted at one hundred posts because the network had shown an exceptional slowdown demonstrated it was being stressed significantly.
The results of the experiment were reasonably predictable, but the stress bot helped illuminate them. The five, twenty posts batches had tps values between 10.72 tps and 17.02 tps. The three, fifty post batches had tps values between 9.89 tps and 15.08 tps. The one hundred post batches handled tps at the lowest 5.32 tps. This demonstrated a significant predicament for the GBYTE network. The GBYTE network is prone to DOS attacks through repetitive posting as it would dramatically slow the entire network.
Test 2 for GBYTE: February 13, 2018
With the first test demonstrating the likelihood a DOS attack would slow the entire network what will the second test show? The second test was slightly different as the goal was not to send a significant number of posts to the DAG network to see how it handles the traffic. Unfortunately, the network did not perform well in the first test, exposing the proneness of the network to DOS attacks through repetitive posting. The second test sought to get the transaction rate figure under sustained posting conditions. This is contrary to the first test where they flooded the network all at once.
The second stress test had numerous batches of fifty units. These units were sent to the network incrementally starting at ten tps eventually arriving at 18 tps, one tps at a time (11, 12, 13, etc.). This approach was dramatically different than the first test which solely sent large numbers of contracts all at once to the network. However, what would the results demonstrate?
The GBYTE network successfully adjusted itself until 15 tps was surpassed. At the stress levels of 16 tps, 17 tps, and 18 tps, the network started becoming noticeably congested. The results were:
- At 15 tps, the response was 14.74 tps.
- At 16 tps the response was 14.32 tps.
- At 17 tps the response was 12.97 tps.
- At 18 tps the response was 10.04 tps.
A small bug was found during the first stress test and the monitoring of confirmation times during the second test had to be delayed until the final test as it had yet to be fixed. The results were clear, following 15 tps the network demonstrated a dramatic decrease in response tps.
The Final Stress Test: February 15, 2018
The final stress test took place on February 15, 2018. The goal of this test was to identify changes in confirmation time (if there were any), during high-stress conditions. The first two tests were mainly concerned with raw tps figures. Varying confirmation times were much less of a concern during Test 1 and 2. The average confirmation time is between 10 and 15 minutes according to daily users. For the third test, the stress bot posted two continuous flows of fifty posts at 15 tps. This was separated by a two-minute resting period followed by another 50 posts. The confirmation times were measured, and the results were the most positive out of all three tests.
The results indicated all posts were simultaneously confirmed less than fifteen minutes after they were sent to the network. This confirms the load on the network has no negative impact on the confirmation times.
Conclusion: Life as a Cryptocurrency is Stressful!
Byteball’s main net seems to be able to handle 17 tps during peak periods and 15 tps during a period of constant flow. When these levels are surpassed, the network shows significant levels of deterioration in speed. These numbers are actually fantastic, placing GBYTE among the fastest functional cryptocurrency networks in production.
Source: Good Audience
The bigger problem is once 17 tps is breached there is exponentially a larger impact on the slowing of the network with each additional post/transaction. If the network is attacked with a flood of posts to slow it intentionally, this will result in a very successful DOS attack. Normal blockchains have tps limited by the block size. GBYTE, including the conditional smart contract feature, makes it so there is an ability to flood the network with spam conditional requests.
GBYTE is not a favorite coin of the King’s until this issue with their network is resolved. However, expect GBYTE to rise in value quickly as the adoption of their conditional payment platform increases and they resolve the likelihood of a DOS attack.
To read the King’s prior articles, to find out which ICOs he currently recommends, or to get in contact directly with the King, you can on Twitter (@JbtheCryptoKing) or Reddit (ICO updates and Daily Reports).