The most recent couple of months have been occupied for our blockchain group. We are gaining awesome ground and need to share what we have been doing with you.
HF 20
Barely a month prior we settled on the extreme choice to put off HF 20 usage for upgrading our application system. We examined this choice with the witnesses and they were steady of this change. We are as yet anticipating actualizing and discharging HF 20 in the coming months.
Scaling the Blockchain
Meanwhile, we have been dealing with some major changes to help scale Steem. We presently run 7 steemd hubs in AWS to help steemit.com and should keep on running increasingly to help our development. The key issue is that Steemd is single strung, in any case, we are well along during the time spent making it multithreaded.
Envision the steemd server program is a stove. At this moment, it has a solitary burner and we have to make a supper. A couple of the dishes are the distributed system, tolerating and assessing pieces, and reacting to API asks for information customers like steemit.com. We are quickly exchanging pots, dish, and skillets onto the single burner to make every part of the feast. The expanded demand volume from steemit.com is postponing the dinner.
Hitherto, we have been purchasing additional single burner stoves to disperse the heap since this was the manner by which blockchains were initially outlined. However, this is not perfect since the greater part of our stoves (our PCs) really have 8 burners (CPU centers) and we are just utilizing 1 each. We have been rolling out improvements to the steemd server program with the goal that we can utilize each of the 8 burners and not require the same number of stoves. That way we can put a skillet on one burner, a pot on another, container on another, you get the thought.
Already, we had a solitary application that took care of the p2p code, the database, modules, and APIs. Presently, everything is a module and just needs to speak with different modules that they specifically rely upon. This expanded level of seclusion takes into consideration faster improvement, simpler code audit, and more noteworthy parallelism. It will likewise enables us to switch parts in and out without breaking the more prominent entirety. We are currently creating with building squares instead of particularly composed perplex pieces.
Texture
Our initial tests are demonstrating a 5-overlap increment in throughput for overhauling API asks. This refactor additionally establishes the framework to scale steemd into the future and actualize the Steem Blockchain Fabric.
There are a few API changes on the steemd hub that should make getting to information in the blockchain less demanding and permit overhauling of our APIs without breaking existing calls. We have included a condenser_api which uncovered all current APIs so we don't break in reverse similarity with existing applications. We have additionally modified the jsonrpc parser and it is totally jsonrpc 2.0 agreeable. We will report when this is conveyed to arranging.
The greater part of the work we have been doing is very specialized and testing, yet we anticipate keeping the group side by side of our work and methodologies, on quite a bit of which we seek the Witnesses and group for coordinated effort. We're upbeat to report that we are certain these arrangements will give a vigorous establishment to meeting the majority of our future scaling needs and empower us to turn our regard for the elements all of you need to see.
Group Steemit
Source- https://steemit.com/steemitdev/@steemitdev/steem-blockchain-update-august-2017
Not indicating that the content you copy/paste is not your original work could be seen as plagiarism.
Some tips to share content and add value:
Repeated plagiarized posts are considered spam. Spam is discouraged by the community, and may result in action from the cheetah bot.
nice post