Today I begin our weekly dev ops updates. We hope to keep our supporters better informed about the updates we are making to improve user experience and our contributions to the network.
This week we spent a lot of time redesigning our infrastructure.
We upgraded our Producing and APi nodes in an effort to reduce our cpu billing time. Our old producing nodes relied too much on turbo boost which didnt seem to work well with the load we see from Nodeos. We upgraded to cpu’s with a much higher base clock and reduced our cpu billing significantly. See the attatched image.
We also redesigned our monitoring to be more reliable. We now hit multiple endpoints to verify our block production and reduced the time it takes for us to be alerted in the case of our nodes going offline.
Lastly we hardened our API nodes and began the updates necessary to reduce our api response times. In order to do this we have began setting up new proxy nodes which are better distributed geographically. We hope to have our api response times down to under 100ms for most users by the time I post this update next week.
Stay tuned for our next update. We will be posting them at 12:01pm CST every friday!