Building an excellent blockchain
A great blockchain can only be built on a solid foundation. On Telos this involves 21 active block producers and 30 standby producers who are required to participate in block production on a regular basis. The fact that 30 standby producers participate in block production means that all active and standby producers impact on the health and performance of the chain. We support block producers who are doing their part in building sound infrastructure required to make Telos the best blockchain. We vote for BPs who are;
- Compliant with the Telos constitution and minimum infrastructure requirements.
- Provide excellent technical services. This includes the best CPU performance, the the fewest missed blocks and public API functionality.
- Geographic diverse.
- Actively participate in growing the adoption of Telos.
Who “teloszaproxy” is voting for
Note : methodology has been updated to include consideration of P2P and public API based of data provided by EOS Metal (for more see here)
A block producer ranking methodology has been developed to yield a balance between CPU benchmark, missed block rates and a functioning public API. This methodology is described below;
- Missed block rates are normalised using the following formula – 1 000 000 x (missed blocks)/(blocks producer + missed blocks). It is necessary to normalise missed block counts to make the measure fair for top 21 producers who produce the majority of blocks and are therefore most likely to miss blocks. Block producers were ranked according to missed block rates.
- Block producers were ranked using AlohaEOS’s CPU benchmark (see AlohaEOS Telos Block Producers Benchmarks).
- A block producer score was calculated as (BP score) = (missed block rate rank) + (CPU benchmark rank). This score was used to rate block producers (lower score is better).
- The public API and P2P connections were checked using the validation tool provided by EOS Metal. Producers with errors were checked manually to verify that the error was valid. Producers with verified errors were excluded from the analysis. Producers with warnings were penalised in the analysis.
- Select 10 producers based on geographic diversity by selecting the top 2 block producers from each of the following regions – North America, Europe, Asia, Central and South America, Africa and Australia – based on BP score from above.
- Select a total of 15 top 21 producers as those that have already been selected through geographic diversity plus the remaining best performers based on BP score.
- Select a total of 15 top standby as those that have already been selected through geographic diversity plus the remaining best performers based on BP score.
This methodology should allow us to select 15 excellent top 21 producers and 15 excellent standby producers while maintaining our goal of geographic diversity.
Our votes based on data collected on 1 March 2019 and API availability on 7 March 2019 are as follows;
|Producer name||Blocks produced||Blocks missed (1)||CPU benchmark||Reason|
|eosbarcelona||565 235||44||1.102||Top 21|
|octagontelos||56 3335||96||1.025||Top 21|
|eosmetaliobp||572 567||105||1.154||Top 21|
|telosmadrid1||564 986||571||1.096||Top 21|
|telosmiamibp||568 429||4||1.386||Top 21|
|caleosblocks||577 188||471||1.182||Top 21|
|votedutcheos||573 501||51||1.493||Top 21|
|telosuknodes||553 747||14||1.584||Top 21|
|21zephyr1111||553 946||72||1.506||Top 21|
|telosglobal1||521 894||697||1.331||Top 21|
|telosgreenbp||548 692||858||1.308||Top 21|
|telosdacnode||562 494||2 935||1.227||Top 21|
|tlsvenezuela||560 335||6 005||1.227||Top 21|
|eosiodetroit||568 980||736||1.587||Top 21|
- Missed block counts normalised to make the measure fair for top 21 producers who produce the majority of blocks and therefore are more likely to miss blocks.
The next table lists the block producers not included in our vote. For some of these producers CPU or block statistics are not available. It was therefore not possible to include them. Lists will be updated on a regular basis. When statistics become available for these producers they will be considered. It must also be emphasised that missed block rates are reported as missed blocks per 1 million blocks. It is necessary to normalise missed blocks in this way to make comparisons fair on the top 21 producers who produce the majority of the blocks and therefore are likely to have the most missed blocks in absolute terms.
|Producer name||Blocks produced||Blocks missed (1)||CPU benchmark|
|goodblocktls||563 335||2 709||1.018|
|blindblocart||532 361||2 660||1.358|
|telosgermany||17 435||2 061||1.449|
|eosvibesbloc||19 816||7 015||1.338|
|bigironbptex||535 150||1 167||1.515|
|eosgermanybp||13 273||90 454||6.381|
|telosafrique||20 613||113 534||7.027|