Trx. Id: 38ea078299d98a9636d6223af96235793c0f99735abf6957cf5609eb31286eac
Expiration Time: 2024-02-24T08:17:51.000
ACTIONS
| CONTRACT | NAME | AUTHORIZATION | DATA |
|---|---|---|---|
| labs.wax | editprop | waxpaca.fx active |
road_map : 45$ / month for each ship node, 6$ / month for each frontend, 10$ / month for load balancer, 30$ / month for each api server. 125$ / month for continous maintenance. And 500$ to create a full api documentation. A total of 6321$ tax included is requested for a year of hosting, maintenance, new infra deployment and documentation creation. Hosting money will be kept in a « infra.we » account and any extra left after a year will be use to prolong hosting, everything will be easily verifiable on chain. proposal_id : 189 mdbody : WaxOnEdge has been such an useful tool for the community. Since previous proposal #111, the work has not stopped, and the platform keep getting new updates. Its usage has been organically increasing without any active promoting and received a lot of good feedback. Currently WaxOnEdge is serving 3.8M requests / month on frontend + 38M requests / month on API with an average of 500 daily unique visitors and 6200 unique visitors on last two weeks (goaccess stats). Recently the current architecture has shown its limit, running on only two servers. If one crash the whole platform is down also it?s impossible to make an update without downtime. And people are starting to run bots for automated swapping + data scrapping. The 3000$ grant given on previous proposal #111 is coming to an end since WoE has been online for more than a year now. And the collected fees are not enough to fully cover servers costs. Aiming to keep WoE a community tool and improve the quality of service, we kindly ask for a new hosting grant. With this proposal WaxOnEdge aims to double its number of servers + add public servers so bots will be able to migrate their requests to those. The architecture will be improved to be behind a HAProxy load balancer + WAF + Rate limiting to prevent abuses. Thanks to the new Wax Leap 5 which introduced the mapped\_private mode we can now run ship node with only 32GB ram instead of 64GB but 128GB recommended before so we will be able to get cheaper servers. The API itself requires 4 good cpu cores (not cheap vcore from a vps) + 10GB ram. We?re planning to deploy 2 ship node server + 1 load balancer + 2 frontend + 3 WaxOnEdge API servers (1 for WoE prod + 1 for public endpoints and 1 for both prod failover + public). The 2 ship nodes are private to be able to get state\_history stream + hammer get\_table\_rows on API init, they will only be accessible from backends and not frontend (most frontend queries are woe api, so public rpc endpoint is not required). The trading history size will be increased, currently it starts from mid december 2022, but we need to evaluate how much is possible since it will increase load on the database. In addition to deploying new servers, an endpoint documentation of the current API will be created so developers will be able to set up their bot requests on top of WaxOnEdge public APIs. image_url : https://waxonedge.app/og_preview.jpg description : Improving infra of WaxOnEdge by renting more servers + creating public endpoints and documentation. title : WaxOnEdge hosting improvement category : infra.tools estimated_time : 1 |