Technical Roundup: stress test results
Over the last months, the dev team’s been busy coding and polishing the key components of the Papyrus ecosystem. The architecture of Papyrus is natively designed for global scaling and high productivity and to implement it, our development process includes regular performance testing to detect potential bottlenecks before they impact system productivity.
Last week, the dev team ran an RTB capacity stress test to identify and eliminate bugs. The stress test was held on the channel layer and consisted of 3 nodes. During the test, the team achieved stable operation within 24 hours handling 1000 ad events / second (e.g., impressions, clicks) on relatively weak hardware (1 Gb RAM, 1 CPU, 100 Gb HDD for each node). Bearing in mind the current weak hardware configuration, it’s impressive results that validate our technological choices. We plan to conduct a full-scale performance testing soon.
Following our own high standards of quality and performance does not prevent us from regularly delivering new functionality and features:
- Auditor integration systems were implemented and tested and are now ready to be added to the MVP.
- All ad events (e.g., impressions, clicks) will now be aggregated and stored to the IPFS storage, and the hash links of these IPFS blocks will be stored in the Ethereum blockchain.
To show results of our work on backend channel nodes, we have improved our web-part. There will be Advertiser and Publisher accounts in the MVP version such that advertisers can easily start campaigns and watch the traffic quality, whereas publishers will be enabled to manage ad inventory effectively.
We are getting closer to the release of the MVP, which is planned for the upcoming weeks. It’s actually even earlier than you may expect. More details to follow — stay with us!