scale.bythebay.io: Martin Zapletal Interview
Recording: scale.bythebay.io: Martin Zapletal Interview
[Music] my name is Martin Zapata I'm a director of engineering at fake solutions bomb tech media company we're building or building large-scale extremely reliable and scalable and elastic solutions so we're running into a lot of problems associated with building these type of systems from the reliability from resilience the cost of infrastructure at that scale to observability of the system in production to making sure make sure that everything works as as it's supposed to there's a variety of issues with building systems of this of this quite extreme scale some of them include the resiliency and reliability of the system so we need to be able to handle things like even catastrophic data center failures but node failures and things like that are extremely common so we have to be able to handle those another problem a common one I think is observability of the system we need to make sure that we understand what the system is doing from both technical and business point of view to make sure that we can we can we can notice and fix any potential issues that happen on production [Music] some of the problems that the industry needs to solve around scaling is again resiliency and make it and being able to build global scale systems that can handle failures and also performance also be able to be performant at that extreme scale while being cost efficient because cost efficiency is another problem that the industry industry has to solve because building these extremely large scale systems becomes expensive the problems are amplified by the scale of the problem so solving that solving the cost efficiency I think is another another really important feature that will have to handle or we're already handling and will be handling more in the future [Music] we use we use functional programming heavily and I see benefits at multiple levels one is the the lower level so referential transparency equation reasoning type safety as well as interest of of engineers but also at the system level so because functional programming provides correctness and safety which is really useful because it handles a part of the problems it solves part of the problems that we we have to solve when building a large-scale systems so that we don't have to worry about that particular area and we can focus on the other things that matter because it solves these problems compile time so we don't run into them during during runtime I really like that the conference brings together people from multiple areas that are sold sort of connected also functional programming machine learning and data science and reactive because they work really closely together although they may not not be realizing it immediately so I like and the other thing I like is there are many interesting talks and conversations about new technologies and and new niche markets and and new upcoming solutions and the future which is really really interesting