APEX Network — Community Supernode Initiative — Experiences
On July 10th APEX Network announced its community supernode initiative, reimbursing rental costs for dedicated servers for selected community members deploying supernodes on our blockchain. The response was overwhelming, and within a few short days we had received dozens of applications from community members who wanted to help us test various aspects of the current blockchain technology.
Through this testing phase we have gained valuable knowledge, and we are now ready to move on to the next phase. Before doing that however, let’s take a look at what has been happening over the last couple of weeks.
The community members signing up to deploy supernodes came with a wide variation of technical skills — from the entry level node operators who had barely used a Linux based operating system before to experienced network and software engineers. In cooperation with our community management team and the devs in Shanghai the number of supernodes rapidly grew, reaching a total of 45 supernodes deployed across 5 continents.
Some notes on the APEX Network DPOS consensus
The way blocks are produced in a DPOS environment is very lightweight on the P2P logic side. Instead of having to send synchronization messages after each block (which would be unfeasible with 0,5 second block rate), each producing supernode gets a fixed time slot for block production which is based on the block production rate and the timestamp of the Genesis block.
On our blockchain the top 21 supernodes by vote count are producing blocks, while the remaining nodes stay synchronized, ready to take on production when voted in to do so. Using DPOS consensus with 0,5 second block time is quite ambitious on a global scale, hence one of the goals of the community supernode initiative was to test the stability and decentralization capability of the APEX Network blockchain.
How did the testing go?
During our testing we have had block producing supernodes spanning five continents (North America, Europe, Africa, Asia and Australia) working in concert without experiencing problems, a fact in which we take pride considering that decentralization is one of the potential weak spots of the DPOS consensus by nature.
Furthermore we note that our community supernodes did not have problems synchronizing, registering and producing blocks on the network. In addition we tested rotating block production among the supernodes through the voting functionality, which also worked without issues. In short; the initiative proves that the APEX Network blockchain can work on a global scale.
Community activity
During the community supernode initative our tech group has seen an exponential increase in activity, and it has been our pleasure to see our community members taking such an active interest in our technological developments. In addition to this, there has been a high level of technical assistance and Q&A about the blockchain through direct messages from interested community members and node operators working on their deployment. With your help we hope to maintain a healthy level of activity in the tech group going forward, both through the continuation of the supernode initative and general technical discussion!
Improvements made and bugs detected
- During initial testing significant improvements were made to synchronization speed which was effectively quintupled (5x) through implementation of a new block synchronization protocol. Further work will be done in this area as mentioned in earlier tech articles.
- A rare issue where synchronization of a new node would slow down and eventually halt was debugged
- Message load was significantly reduced through INV message handling improvements, and further optimizations will be done
- A bug was found (msg.value) in the VM, this has already been rectified in the next iteration of the code
- A bug which could cause block production to be interrupted was identified and fixed
What’s next?
During the next phase of the community supernode initative, we will among other things be testing the hard fork functionality of the network. Hard forks are needed in the case of major bugs or feature changes in the core code where the version of the code needed to be implemented is not backwards compatible, which makes this functionality crucial for future proofing the blockchain.
During this phase a limited subset of technically proficient community supernoders will be asked to continue operating their nodes, and they will be offered reimbursement of ongoing server costs. These node operators have been chosen by the team based on technical proficiency and performance during the initial part of the community supernode initiative. Those asked to participate will be expected to be able to handle the technical side with larger autonomy. Technical assistance will of course still be provided, but entry level linux support will be limited.
All the chosen candidates will receive an email requesting their ongoing assistance by 06.08.2019 22:00 UTC+8. All other current node operators are free to continue participation in the program, but will be required to cover their own hardware rental costs. As previously stated, please make sure to cancel your server subscriptions in due time to avoid unnecessary costs if you should not wish to maintain your supernode beyond the month of reimbursement. We would like to extend our most sincere gratitude to everyone who have assisted us in the testing so far!
Until next time,
APEX Team
APEX Network
- Visit our Website
- Follow us on Twitter
- Join our Telegram Community
- Subscribe to our Announcements Channel
- Subscribe to our Subreddit
