�ݺ�ߣshows by User: artgon

�ݺ�ߣshows by User: artgon / http://www.slideshare.net/images/logo.gif �ݺ�ߣshows by User: artgon / Wed, 24 Oct 2018 23:15:01 GMT �ݺ�ߣShare feed for �ݺ�ߣshows by User: artgon Scalable Anomaly Detection (with Zero Machine Learning) /slideshow/scalable-anomaly-detection-with-zero-machine-learning/120618120 sl2018-raju-181024231501
In a large scale distributed system, detecting and pinpointing failures gets exponentially harder as an architecture gets more complex. Netflix's cloud architecture is composed of thousands of services and hundreds of thousands of VMs and containers. Failures can happen at any level and can often cascade quickly, some can cause massive outages on several systems, while others only only break one or two. This creates a needle in a haystack problem that requires automated and precise detection. Zuul, as the front door for all of Netflix's cloud traffic, sees all requests and responses and is ideally positioned to identify and isolate only the broken paths in the maze of microservices. We leveraged Zuul to stream real-time events for each request-response and built an anomaly detector to automatically identify and alert services in trouble. We scaled this detector to thousands of nodes, handling millions of requests, without a single line of machine learning. Sometimes you need machine learning and sometimes you don't. Although it's en vogue to apply machine learning to every problem, it can be more practical and approachable to solve certain problems with old-fashioned math! In this talk, we'll discuss how we built this system with stream processing, anomaly detection algorithms, and a rules engine. We will also deep-dive into the anomaly detection algorithm and show how sometimes a simple, elegant algorithm can be just as good as any sophisticated machine learning.]]>
In a large scale distributed system, detecting and pinpointing failures gets exponentially harder as an architecture gets more complex. Netflix's cloud architecture is composed of thousands of services and hundreds of thousands of VMs and containers. Failures can happen at any level and can often cascade quickly, some can cause massive outages on several systems, while others only only break one or two. This creates a needle in a haystack problem that requires automated and precise detection. Zuul, as the front door for all of Netflix's cloud traffic, sees all requests and responses and is ideally positioned to identify and isolate only the broken paths in the maze of microservices. We leveraged Zuul to stream real-time events for each request-response and built an anomaly detector to automatically identify and alert services in trouble. We scaled this detector to thousands of nodes, handling millions of requests, without a single line of machine learning. Sometimes you need machine learning and sometimes you don't. Although it's en vogue to apply machine learning to every problem, it can be more practical and approachable to solve certain problems with old-fashioned math! In this talk, we'll discuss how we built this system with stream processing, anomaly detection algorithms, and a rules engine. We will also deep-dive into the anomaly detection algorithm and show how sometimes a simple, elegant algorithm can be just as good as any sophisticated machine learning.]]> Wed, 24 Oct 2018 23:15:01 GMT /slideshow/scalable-anomaly-detection-with-zero-machine-learning/120618120 artgon@slideshare.net(artgon) Scalable Anomaly Detection (with Zero Machine Learning) artgon In a large scale distributed system, detecting and pinpointing failures gets exponentially harder as an architecture gets more complex. Netflix's cloud architecture is composed of thousands of services and hundreds of thousands of VMs and containers. Failures can happen at any level and can often cascade quickly, some can cause massive outages on several systems, while others only only break one or two. This creates a needle in a haystack problem that requires automated and precise detection. Zuul, as the front door for all of Netflix's cloud traffic, sees all requests and responses and is ideally positioned to identify and isolate only the broken paths in the maze of microservices. We leveraged Zuul to stream real-time events for each request-response and built an anomaly detector to automatically identify and alert services in trouble. We scaled this detector to thousands of nodes, handling millions of requests, without a single line of machine learning. Sometimes you need machine learning and sometimes you don't. Although it's en vogue to apply machine learning to every problem, it can be more practical and approachable to solve certain problems with old-fashioned math! In this talk, we'll discuss how we built this system with stream processing, anomaly detection algorithms, and a rules engine. We will also deep-dive into the anomaly detection algorithm and show how sometimes a simple, elegant algorithm can be just as good as any sophisticated machine learning. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/sl2018-raju-181024231501-thumbnail.jpg?width=120&height=120&fit=bounds" /><br> In a large scale distributed system, detecting and pinpointing failures gets exponentially harder as an architecture gets more complex. Netflix's cloud architecture is composed of thousands of services and hundreds of thousands of VMs and containers. Failures can happen at any level and can often cascade quickly, some can cause massive outages on several systems, while others only only break one or two. This creates a needle in a haystack problem that requires automated and precise detection. Zuul, as the front door for all of Netflix's cloud traffic, sees all requests and responses and is ideally positioned to identify and isolate only the broken paths in the maze of microservices. We leveraged Zuul to stream real-time events for each request-response and built an anomaly detector to automatically identify and alert services in trouble. We scaled this detector to thousands of nodes, handling millions of requests, without a single line of machine learning. Sometimes you need machine learning and sometimes you don't. Although it's en vogue to apply machine learning to every problem, it can be more practical and approachable to solve certain problems with old-fashioned math! In this talk, we'll discuss how we built this system with stream processing, anomaly detection algorithms, and a rules engine. We will also deep-dive into the anomaly detection algorithm and show how sometimes a simple, elegant algorithm can be just as good as any sophisticated machine learning.

Scalable Anomaly Detection (with Zero Machine Learning) from Arthur Gonigberg

]]> 482 1 https://cdn.slidesharecdn.com/ss_thumbnails/sl2018-raju-181024231501-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Zuul's Journey to Non-Blocking /slideshow/zuuls-journey-to-nonblocking/80385664 arthurgonigberg-zuulsjourneytononblocking-171003020719
Zuul 2.0 is the latest iteration of the gateway application fronting Netflix's API and underlying microservices. It was borne of a need to handle an ever-growing amount of traffic and a similarly ever-growing number of microservices to front. We largely rebuilt Zuul from the ground up, leveraging the Netty framework for its high-performance, non-blocking, event-loop architecture and combined it with RxJava interfaces for a simpler programming model in our filters. Despite our lofty performance expectations for this project we ended up with some mixed results and can definitively say that going asynchronous, non-blocking is not a panacea. This talk will be a deep dive into how we progressively refactored and rebuilt Zuul from a blocking Tomcat application to a non-blocking Netty application and the results we have seen from running it in production over the last year. Specifically, I will review the journey of combining the RxJava and Netty frameworks in rebuilding Zuul, discuss how to make the decision on whether your systems need to be non-blocking, and provide the good and bad with some real-world scenarios.]]>
Zuul 2.0 is the latest iteration of the gateway application fronting Netflix's API and underlying microservices. It was borne of a need to handle an ever-growing amount of traffic and a similarly ever-growing number of microservices to front. We largely rebuilt Zuul from the ground up, leveraging the Netty framework for its high-performance, non-blocking, event-loop architecture and combined it with RxJava interfaces for a simpler programming model in our filters. Despite our lofty performance expectations for this project we ended up with some mixed results and can definitively say that going asynchronous, non-blocking is not a panacea. This talk will be a deep dive into how we progressively refactored and rebuilt Zuul from a blocking Tomcat application to a non-blocking Netty application and the results we have seen from running it in production over the last year. Specifically, I will review the journey of combining the RxJava and Netty frameworks in rebuilding Zuul, discuss how to make the decision on whether your systems need to be non-blocking, and provide the good and bad with some real-world scenarios.]]> Tue, 03 Oct 2017 02:07:19 GMT /slideshow/zuuls-journey-to-nonblocking/80385664 artgon@slideshare.net(artgon) Zuul's Journey to Non-Blocking artgon Zuul 2.0 is the latest iteration of the gateway application fronting Netflix's API and underlying microservices. It was borne of a need to handle an ever-growing amount of traffic and a similarly ever-growing number of microservices to front. We largely rebuilt Zuul from the ground up, leveraging the Netty framework for its high-performance, non-blocking, event-loop architecture and combined it with RxJava interfaces for a simpler programming model in our filters. Despite our lofty performance expectations for this project we ended up with some mixed results and can definitively say that going asynchronous, non-blocking is not a panacea. This talk will be a deep dive into how we progressively refactored and rebuilt Zuul from a blocking Tomcat application to a non-blocking Netty application and the results we have seen from running it in production over the last year. Specifically, I will review the journey of combining the RxJava and Netty frameworks in rebuilding Zuul, discuss how to make the decision on whether your systems need to be non-blocking, and provide the good and bad with some real-world scenarios. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/arthurgonigberg-zuulsjourneytononblocking-171003020719-thumbnail.jpg?width=120&height=120&fit=bounds" /><br> Zuul 2.0 is the latest iteration of the gateway application fronting Netflix's API and underlying microservices. It was borne of a need to handle an ever-growing amount of traffic and a similarly ever-growing number of microservices to front. We largely rebuilt Zuul from the ground up, leveraging the Netty framework for its high-performance, non-blocking, event-loop architecture and combined it with RxJava interfaces for a simpler programming model in our filters. Despite our lofty performance expectations for this project we ended up with some mixed results and can definitively say that going asynchronous, non-blocking is not a panacea. This talk will be a deep dive into how we progressively refactored and rebuilt Zuul from a blocking Tomcat application to a non-blocking Netty application and the results we have seen from running it in production over the last year. Specifically, I will review the journey of combining the RxJava and Netty frameworks in rebuilding Zuul, discuss how to make the decision on whether your systems need to be non-blocking, and provide the good and bad with some real-world scenarios.

Zuul's Journey to Non-Blocking from Arthur Gonigberg

]]> 1315 2 https://cdn.slidesharecdn.com/ss_thumbnails/arthurgonigberg-zuulsjourneytononblocking-171003020719-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Scala meetup - Objectify /slideshow/scala-meetup-objectify-15072182/15072182 scala-meetup-objectify-121107143347-phpapp01
]]>
]]> Wed, 07 Nov 2012 14:33:44 GMT /slideshow/scala-meetup-objectify-15072182/15072182 artgon@slideshare.net(artgon) Scala meetup - Objectify artgon <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/scala-meetup-objectify-121107143347-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds" /><br>

Scala meetup - Objectify from Arthur Gonigberg

]]> 815 2 https://cdn.slidesharecdn.com/ss_thumbnails/scala-meetup-objectify-121107143347-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 https://cdn.slidesharecdn.com/profile-photo-artgon-48x48.jpg?cb=1540422796 https://cdn.slidesharecdn.com/ss_thumbnails/sl2018-raju-181024231501-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/scalable-anomaly-detection-with-zero-machine-learning/120618120 Scalable Anomaly Detec... https://cdn.slidesharecdn.com/ss_thumbnails/arthurgonigberg-zuulsjourneytononblocking-171003020719-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/zuuls-journey-to-nonblocking/80385664 Zuul's Journey to Non-... https://cdn.slidesharecdn.com/ss_thumbnails/scala-meetup-objectify-121107143347-phpapp01-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/scala-meetup-objectify-15072182/15072182 Scala meetup - Objectify