Stacktical

A scalability report for every new release of your product

The easiest and fastest way to know how many simultaneous users your online products can handle.

Reviews

Discussion

You need to become a Contributor to join the discussion - Find out how.
Nicolas Grenié
Nicolas GreniéHunterHiring@picsoung · Developer Advocate, Typeform
Hello Hunters :) Today I am really pleased to share with you Stacktical, an app we could all use. Who never suffered the "ProductHunt effect" (or "Hacker News") on their website? Who never saw a sudden huge spike of traffic after a TV show talked about your startup? Remember how embarrassing it is to serves 404 or 503 pages to people coming to check you out? Well, this should not happen, if you run Stacktical before any major Marketing action or product release. They will analyze your stack, and tell you how much you can handle and which actions you should take to scale better. So give it a try and use it on your next project ;)
Wilhem Pujar
Wilhem PujarMakerHiring@wilhempujar1 · CEO at Stacktical
@picsoung Thanks a tons for the hunt Nicolas! Hello Product Hunt family :) tl;dr: You are not getting the better bang for your hosting buck if you are not shipping scalable softwares! We had the idea of Stacktical on a french election day. At the time I was helping the country's biggest broadcast TV channel with the web architecture of two new products they were about to launch. You would think that on election day, they were somehow expecting a bit of traffic to their news website, right? Well things still went south pretty quickly 🚨 ! And the infrastructure engineering team spent all night working on experimental scaling operations that would restore service and help them cope with the traffic. That's exactly when I told to myself that if you have a clear understanding of how your online products scale at all time, you'd have the tools to prevent such situation from happening in the first place. At the very least you would be better prepared to handle it. Predictive Analytics and AI were booming at the time, reinforcing this feeling that there definitely had to be a better way to ship products while ensuring their functioning at scale... So we started working on just that. At the beginning of this journey, we quickly realised that before Stacktical, there was no testing product dedicated to scalability, and no scalability testing practices that could keep pace with the speed of modern, continuous developments. As a consequence, companies would just throw money at their scalability problems. This is getting worse with the rise of Cloud Computing and ways to "focus on building great products" that are in fact significantly increasing the waste of computing power, energy and money of your hosting infrastructure. We need to stop this waste. Stacktical is the first predictive platform dedicated to continuous scalability testing, reporting and tuning. A super fast and effortless way to predict how many max simultaneous users your new release can handle before it is shipped to production, and share insights with all scalability stakeholders in your company. So DevOps engineers, Developers, Marketers, Managers can finally answer this simple question: Will it scale? Talking about questions, happy to answer any question you might have!
Nicolas Vérité, Nÿco
Nicolas Vérité, Nÿco@nyconyco · ProductOwner-LeanStartup-DesignThinking
@picsoung @wilhempujar Well done Wilhem, congrats! And thanks Nicolas for sharing this. Wilhem, I'd like to know more about Stacktical! It seems we are close, both geographically in Paris, and also in the scalability arena. We deliver a massively scalable real-time messaging/chat platform, for businesses to build social chat apps. We have built our own Continuous Load Testing platform. Does Stacktical only focus on the web? What technology stacks can it handle? Can it support Erlang/Elixir and other BEAM-based systems?
Wilhem Pujar
Wilhem PujarMakerHiring@wilhempujar1 · CEO at Stacktical
@nyconyco Hey Nicolas, thanks for the support and the kind words! The Stacktical predictive engine turns very small samples of load testing results (a dozen or so) into accurate and actionable scalability insights (so you don't have to spend all day load testing and do the interpretation yourself). If you already have a Continuous Load Testing platform (which is an excellent CI/CD practice we'd love to see more), you just need to input your load test results to Stacktical. For now, you'll have to manually input them but we're already working on a way to run a scalability test using existing load testing results. ETA next week! In its current version, our Docker plugin application at https://hub.docker.com/r/stackti... runs its own set of load tests before a scalability test to minimize friction for less mature users (understand without a "Continuous Load Testing" platform). Basically, you'll be soon able to plug your Continuous Load Testing platform to our Continuous Scalability Testing platform and generate a scalability report with every new load test in your pipeline. We'll also be providing access to our API at some point. Just need to properly document the beast :)
Oh wow this is amazing. Nice job @wilhempujar! How are the reports generated? Do you have recommendations for every stack? How do you specialize recommendations?
Wilhem Pujar
Wilhem PujarMakerHiring@wilhempujar1 · CEO at Stacktical
@francium Hey! Thanks for the kind words Francesca! We're using a combination of advanced analytics techniques and AI to generate the reports from (a dozen) load testing results. Load testing results are required to run our predictive engine. But we're providing a plugin for those who don't have any though! As I told @nyconyco our Docker plugin at https://hub.docker.com/r/stackti... does both a load test AND a scalability testing in one shot. The beauty of Maths is that our platform works with pretty much every measurable stacks and services. You can test a front-end service, an API, a database cluster behind an IP address and pretty much everything that can be described as "users versus operations per seconds", which is typical of load testing. Our plugin only supports http and https for now though. But we are working on ways to lift this limitation. We're very much interested in providing (more) user friendly recommendations that are tailored to the nature of the tested service. I'll have more to share about that soon enough :) Hope that answers your questions!