Steem SMT Ideas: Full P2P USER Case - Post #46
Illustrated below is the Proposed Pure P2P Setup:
FLOW OF DATA EXCHANGE FROM USER TO THE BLOCKCHAIN
It will be of a good benefit if you will read my previous post #45 as it is related to this post, link below.
Imagine the black dot in the middle of this diagram to be the USER (or the DDOS attacker!) wanting to acquire (or stop the flow!) of some blockchain data.
The USER wanting to acquire some blockchain data “always” needs to go thru the levels as arranged in ascending order of levels. The green zone seed nodes (servers) also acts as database servers with pointers as to where and what servers hold the data as inquired by the USER. The database service as pointers is not limited to the green zone however, all seed servers should also be database servers, complete with all pointer data.
Every server as queried should have a timeout set so that the query will go to the next higher level until an answer can be received by the USER.
In summary of the USER’s query flow:
-any server on the green zone queried firstly for the location of the data and the data itself
-if no reply within a specific period the query goes to any server on the red zone
-and so on going up the circumference ladder, until an answer is achieved
Every server is both a database server holding pointer data and a seed node holding the blockchain data.
--
THE DATABASE POINTERS EXPLAINED
As data on the blockchain will be distributed among many seed nodes (not anymore whole blockchain record per one seed node), a pointer as to what server holds what records should be established.
A pointer is not anymore a heavy data in the first place, it just holds certain and few records as far as the database is concerned. This can reside in the memory.
The blockchain data itself can be in the disks even if it is being distributed among many servers. The future holds so much data, and even if we can not clearly see it now, blockchain data will surely grow exponentially over the years when blockchain technology will be finally accepted by the masses and all governments. Having described these, it is just proper that starting this early the blockchain needs to be in a container that can handle huge file sizes as well as distributable. Hard disks are designed for this purpose.
The combination of the database pointer (in memory) and the distributed blockchain data (in disks) will make this approach “limitless” as far as the future of the STEEMD service is concerned.
--
THIS IS NATURALLY LIMITLESS AND TECHNOLOGICALLY SOUND
The number of servers in this approach will surely grow which should be the mindset in the first place when platform upgrades are concerned. Only this time, the servers are sliced thin as the layer of upgrade expands out from the green zone.
Imagine a beehive growing, with pentagon shaped architecture supporting its whole structure.
Nature at its best, applied to technology as the STEEMD service.
--
Previous Post Link:
https://steemit.com/witness-category/@lightingmacsteem/68lq5m-everything-under-the-stars