Failure is Hard

Couple of days ago there was another massive outage on Amazon Web Services. This time was around AWS DynamoDB and that create lots of noise on the the web.

Was not the first time a outage happened and wont be the last time for sure. There are some summary report from aws you can check.

So is this the end of cloud computing? Should we stop using cloud as single provider? All this questions start emerge from the web. I dont think thats the case - for sure there are some lessons learned from this last incident. The reality is distributed computing is hard and will always be.

Distributed Systems Fallacies - The network is reliable - Fail

You are doing distributed computing failure will happen. The Cloud is not magical. There are frameworks and approaches we need embrace more to being more anti-fragile. Like having user experience degradation and failure contention. But even if we test for failure will be cases that our tests wont catch. We need just make sure we keep learning from failures but more distribution is not the answer - i MEAN having multiple cloud providers is actually more problems because this will require more integration work.

Your Data Center 

One thing some companies do(Often when they have BIG money and the expertise) it`s run your own data center. IF you do have the hardware and expertise you can go for it. You might have better availability because is just your stuff there but this dont make the task easy and IMHO you still gonna have failure and could be even worst them use AWS or AZURE.

Not being LOCKED by amazon is or Microsoft is great but carrying this task is not easy and most of IT companies cant do at the same level of expertise.

Multi-Region Architecture - COST / Complexity

Yes we need Multi-Region architecture - we need be able to work with the active-active pattern but to be clear this is not easy at all - this introduce lots of COSTS and COMPLEXITY not to mention LATENCY, REPLICATION ISSUES and this is not easy. Sometimes is necessary yes it is. Do we need for every business i dont think so, Do we gonna need more and more? Absolutely.  This is not a FREE LUNCH there some solution that deal with this better and still lots of extra engineering work.

Takes time to mature it and remember there are 3 other forces happen at some time. 1 is new features, 2 is new software updates and 3 is data migration all this things together on a multi-region arch creates lots and lots of headaches.

Uber and other have they own data center - Most of companies or most of startups cant afford this because is very expensive. The thing is that public cloud like AWS and AZURE are like celebrities and everything that happens with they are on spot lights and are potentiality for good and bad. This means you gonna have failure in your data center but since is just you there is less viral tham is amazon goes down.

The Steps Toward Evolution

IT operations are hard. Failure is even harder. There is one thing on the DevOps Community called Blameless Incident Reviews (Post Mort-ens) and from Agile Coaching and Have Retrospectives and Lessons Learned and its crucial todo this exercises and make changes in the tools and practices we have to build and operate software to make it less anti-fragile.

Failure in Hours per Cloud Providers 


This data is for 2014 so as you can see everybody fails - amazon pretty much one time per year at least but things are getting better and we need continue investing and learning new ways do build operate and test our services. 

Failure will happen - maybe we should just accept it and save some money for when it happens? Now if you decide do in house data center just because of failure this is wrong and based on FEAR and you know what? you will have failure in your company data center too. It`s okay to have your own data center if you have the right reasons, people and money.'

Update 24/09/2015: From 23 and 24 Sep i was in Santa Clara, CA for the 2015 Cassandra Summit and i talk with some Netflix guys i they told me Netflix did not went down during the DynamoDB issue in amazon - They said Netflix went down just on the same data center amazon had issues but the other data centers took over in the active-active pattern - they just operate with 50% capacity.

Cheers,
Diego Pacheco

Popular posts from this blog

C Unit Testing with Check

Image
When I start on IT, several years ago(+15), C was the first language that I learned. But was not the first language I mastered and deployed in production which was...surprisingly Visual Basic 6.0. However, when I learned C or any language back on that old and arcane time noddy was into testing. However now is impossible to work professionally in any language without tests. C is a low-level language but still kicks ass and has several interesting libs/frameworks. People use to fear C a lot because of lack of syntactical sugar but once you get used to it is not that hard at all. Today I want to show how you can do Unit Testing with C. I will not show how to do Unit Testing with C++ but with C ANSI actually. We will be using Check a unit test lib for C ANSI.
read more »

Having fun with Zig Language

Image
Zig is a general-purpose language created in 2016 by Andrew Kelly .  Zig aims to fix the issues C language has while being pragmatic and simple at the same time. Zig really shines as being simple by not having any hidden control flows, not having hidden memory allocations, no pre-processors, and no macros. Basically, it's one language and one language only. Since pre-processors and macros can get really complex and become their own universe. Sure it's not the Zig Philosophy.  Zig Compiler(Zig cc) also is a C compiler, and the interoperability from C to Zig is very good. Clearly, we can see influence from other languages like Go and Rust; however, Zig is pretty unique and different from Go and Rust.  Zig is fast. Zig is perfect for system programming. I bet we will see a lot of solutions being built in Zig. As a languages aficionado, I like to learn at least one language per year and sometimes more; Zig was my 2021 pick, did not regret it. Even if you dont use all languages in p
Keep reading

HMAC in Java

Image
In 2018 AWS CTO(Vogels) said : "Security is everyone's job". For the last 2 years, there was so many famous and big data leaks and breaks that made that statement be very true more than ever. Security often could mean worst performance and worst user experience so in order to get it right you really need to think about the designs before jumping into to code and consider performance and user experience has main requirements.  I was thinking about writing about how to do HMAC in Java for a while and recently Redis 6.0.0 come out and to my surprise, there was a refactoring on the password part in order to use HMAC.  Security easily could scare engineers often because it is not something well spread yet but I believe this will change soon. It does not matter if you have to deal with PII Data or not, security is super relevant because everybody is running their workloads at the cloud or with IoT and Edge devices which means more distribution, more code, more points of fail
read more »