Alternatives to Istio

We just launched the Segment Config API (try it out for yourself here) — a set of public REST APIs that enable you to manage your Segment configuration. Behind the scenes the Config API is built with Go , GRPC and Envoy.

At Segment, we build new services in Go by default. The language is simple so new team members quickly ramp up on a codebase. The tool chain is fast so developers get immediate feedback when they break code, tests or integrations with other systems. The runtime is fast so it performs great at scale.

For the newest round of APIs we adopted the GRPC service #framework.

The Protocol Buffer service definition language makes it easy to design type-safe and consistent APIs, thanks to ecosystem tools like the Google API Design Guide for API standards, uber/prototool for formatting and linting .protos and lyft/protoc-gen-validate for defining field validations, and grpc-gateway for defining REST mapping.

With a well designed .proto, its easy to generate a Go server interface and a TypeScript client, providing type-safe RPC between languages.

For the API gateway and RPC we adopted the Envoy service proxy.

The internet-facing segmentapis.com endpoint is an Envoy front proxy that rate-limits and authenticates every request. It then transcodes a #REST / #JSON request to an upstream GRPC request. The upstream GRPC servers are running an Envoy sidecar configured for Datadog stats.

The result is API #security , #reliability and consistent #observability through Envoy configuration, not code.

We experimented with Swagger service definitions, but the spec is sprawling and the generated clients and server stubs leave a lot to be desired. GRPC and .proto and the Go implementation feels better designed and implemented. Thanks to the GRPC tooling and ecosystem you can generate Swagger from .protos, but it’s effectively impossible to go the other way.

At uSwitch we wanted a way to load balance between our multiple Kubernetes clusters in AWS to give us added redundancy. We already had ingresses defined for all our applications so we wanted to build on top of that, instead of creating a new system that would require our various teams to change code/config etc.

Envoy seemed to tick a lot of boxes:

• Loadbalancing capabilities right out of the box: health checks, circuit breaking, retries etc.
• Tracing and prometheus metrics support
• Lightweight
• Good community support

This was all good but what really sold us was the api that supported dynamic configuration. This would allow us to dynamically configure envoy to route to ingresses and clusters as they were created or destroyed.

To do this we built a tool called Yggdrasil using their Go sdk. Yggdrasil effectively just creates envoy configuration from Kubernetes ingress objects, so you point Yggdrasil at your kube clusters, it generates config from the ingresses and then envoy can loadbalance between your clusters for you. This is all done dynamically so as soon as new ingress is created the envoy nodes get updated with the new config. Importantly this all worked with what we already had, no need to create new config for every application, we just put this on top of it.

How Uber developed the open source, end-to-end distributed tracing Jaeger , now a CNCF project:

Distributed tracing is quickly becoming a must-have component in the tools that organizations use to monitor their complex, microservice-based architectures. At Uber, our open source distributed tracing system Jaeger saw large-scale internal adoption throughout 2016, integrated into hundreds of microservices and now recording thousands of traces every second.

Here is the story of how we got here, from investigating off-the-shelf solutions like Zipkin, to why we switched from pull to push architecture, and how distributed tracing will continue to evolve:

https://eng.uber.com/distributed-tracing/

(GitHub Pages : https://www.jaegertracing.io/, GitHub: https://github.com/jaegertracing/jaeger)

Bindings/Operator: Python Java Node.js Go C++ Kubernetes JavaScript OpenShift C# Apache Spark

To provide employees with the critical need of interactive querying, we’ve worked with Presto, an open-source distributed SQL query engine, over the years. Operating Presto at Pinterest’s scale has involved resolving quite a few challenges like, supporting deeply nested and huge thrift schemas, slow/ bad worker detection and remediation, auto-scaling cluster, graceful cluster shutdown and impersonation support for ldap authenticator.

Our infrastructure is built on top of Amazon EC2 and we leverage Amazon S3 for storing our data. This separates compute and storage layers, and allows multiple compute clusters to share the S3 data.

We have hundreds of petabytes of data and tens of thousands of Apache Hive tables. Our Presto clusters are comprised of a fleet of 450 r4.8xl EC2 instances. Presto clusters together have over 100 TBs of memory and 14K vcpu cores. Within Pinterest, we have close to more than 1,000 monthly active users (out of total 1,600+ Pinterest employees) using Presto, who run about 400K queries on these clusters per month.

Each query submitted to Presto cluster is logged to a Kafka topic via Singer. Singer is a logging agent built at Pinterest and we talked about it in a previous post. Each query is logged when it is submitted and when it finishes. When a Presto cluster crashes, we will have query submitted events without corresponding query finished events. These events enable us to capture the effect of cluster crashes over time.

Each Presto cluster at Pinterest has workers on a mix of dedicated AWS EC2 instances and Kubernetes pods. Kubernetes platform provides us with the capability to add and remove workers from a Presto cluster very quickly. The best-case latency on bringing up a new worker on Kubernetes is less than a minute. However, when the Kubernetes cluster itself is out of resources and needs to scale up, it can take up to ten minutes. Some other advantages of deploying on Kubernetes platform is that our Presto deployment becomes agnostic of cloud vendor, instance types, OS, etc.

#BigData #AWS #DataScience #DataEngineering

Hello :) We are using Datadog on Kong to monitor the metrics and analytics.

We feel that the cost associated with Datadog is high in terms of custom metrics and indexations. So, we planned to find an alternative for Datadog and we are looking into Grafana implementation with kong.

Will the shift from Datadog to Grafana be a wise move and flawless?

As for the new support of service mesh pattern by Kong, I wonder how does it compare to Istio?

Amazon API Gateway vs Apigee. How do they compare as an API Gateway? What is the equivalent functionality, similarities, and differences moving from Apigee API GW to AWS API GW?

As we've evolved or added additional infrastructure to our stack, we've biased towards managed services. Most new backing stores are Amazon RDS instances now. We do use self-managed PostgreSQL with TimescaleDB for time-series data—this is made HA with the use of Patroni and Consul.

We also use managed Amazon ElastiCache instances instead of spinning up Amazon EC2 instances to run Redis workloads, as well as shifting to Amazon Kinesis instead of Kafka.

Postmates built a tool called Bazaar that helps onboard new partners and handles several routine tasks, like nightly emails to merchants alerting them about items that are out of stock.

Since they ran Bazaar across multiple instances, the team needed to avoid sending multiple emails to their partners by obtaining lock across multiple hosts. To solve their challenge, they created and open sourced ConsulMutEx, and an Elixir module for acquiring and releasing locks with Consul and other backends.

It works with Consul’s KV store, as well as other backends, including ets, Erlang’s in-memory database.