Design for Errors

Errors in software are to be expected. Since you expect error you should make design decisions to optimize the likelihood that:

1. Errors are not introduced.
• Use code reviews to catch functional errors.
• Use static analysis to catch coding errors.
2. Errors that are introduced are easily detected.
• Ensure that your code is setup log unexpected errors.
• Use tools like AppDynamics to view the ranking of errors over different time frames.
• Use tools, such as, Splunk to graph errors by type over time. Especially, critical to do this before and after deploying to highlight new errors introduced by the latest deployment.
3. Errors that remain in the software after deployment are either noncritical or are compensated for during execution so that the error does not cause a disaster.
• This is all about testing that things work in a way that you want when they are broken. Here's some examples:
1. Database is down.
   Don't let your service return anything that makes clients think that their order was saved. You have to make the call based on your business if it is better to return error responses or just refuse to accept requests all together. Which of those could get you into more trouble? But don't just code for that. Actually kill your database in the test environment and test it. Setup a functional test that does this every time you build.
2. Disk is full and you can no longer log.
   Same thing as above. Setup a functional test that points the logs at a very small virtual drive and make this test happen with every build.
3. System encoding changed.
   Seriously, this is a thing. On Macs the default is utf-8, on Windows it is ascii, on Linux it is something. I have some Python 3 code that I originally wrote on a Mac but when I moved it to a Windows machine it choked on the encoding all because I didn't set an encoding. I was [unknowingly] depending on the system default. Servers are group property. Maybe only your OPS team can touch them but you probably have more than one Operations Engineer. Or your servers are configured via Chef or something else and everyone misses that single line where someone decided it would be good to explicitly define the system encoding. Boom! Code that was working fails for what most would consider a trivial bit of code.

Such robustness is not easy to incorporate into a design. Some of the ideas that help include the following:

1. Never fall out of a case statement. For example, if there are four possible values for a variable, don't check just for three and assume that the fourth is the only remaining possibility. Instead, assume the impossible; check for the fourth value and trap the error condition early.
2. Predict as many "impossible" conditions that you can and develop strategies for recovery.
3. To eliminate conditions that may cause disasters, do fault tree analysis for predictable unsafe conditions.

Tourism - Family 2016

Home - I've always loved it. I think it is pretty great and that it has a lot of potential to be better.

Design Principles

Design is the set of activities including:

1. defining an architecture for the software that satisfies the requirements
2. specifying an algorithm for each software component in the architecture

The architecture includes a specification of all the building blocks of the software, how they interface with each other, how they are composed of one another, and how copies of components are instantiated (made in memory and executed) and destroyed. The final product of design is a design specification.

How to Ease Growing Pains

As companies get bigger, even child companies (brands, if you prefer), there comes a time when the startup mentality might not be helpful. Here are five ways to make dealing with success easier.

1. Recognize that management is a skill.
   • As you have more development teams, it becomes harder to focus everyone toward the same goal. Good management does that.
2. Hire a COO.
   • It makes sense for young companies to emphasize product development and sales over operations. However, there comes a point when it is more important to hone operations and remove any and all sloppiness.
   • “Customers have to be sure you can deliver.” – Robert Shar
   • Midsized Companies Can’t Afford Operational Glitches
3. Get comfortable saying no.
   • Don’t take on every project that comes your way. Really think about the odds of success. Focus on what is important to your business.
4. Be focused on always getting better at what you already do.
   • New projects can be an unreliable route to growth.
   • “You have to get out of that mode of doing everything differently and turn to repetitive sets of tasks performed by people that become more and more expert in them.” – Derek Lidow
5. Understand that it is not do or die.
   • If you lack clear strategy then there is nothing wrong with focusing on what is already working for you.

The above is a ripoff from your-awkward-phase-and-why-you-should-love-it.

The Paradox of Success

The “paradox of success” can be summed up in four phases:

1. When we have clarity of purpose, it enables us to succeed at our endeavor.
2. When we have success, we gain a reputation as a “go to” person. We are presented with increased options and opportunities.
3. When we have increased options and opportunities, we have increased demands upon our time and energies, it leads to diffused efforts.
4. We become distracted from what would otherwise be our highest level of contribution. The effect of our success has been to undermine the very clarity that led to our success in the first place.

The pursuit of success can be a catalyst for failure.
===
Success can distract us from focusing on the essential things that produce success in the first place.

Software Architects Need Feedback

"Typically, [architects] are farmed out to various projects as consultants, with the aim of ensuring that the project takes off on the right track and avoids mistakes it might make without the wisdom of the consultants. The intent of this practice is laudable, but the outcome is usually sobering: because the consultants are so valuable, having given their advice, they are moved to the next project long before implementation is finished, let alone testing and delivery. By the time the consultants have moved on, any problems with their earlier sage advice are no longer their problems, but the problems of a project they have long since left behind. In other words, the consultants never get to live through the consequences of their own design decisions, which is a perfect way to breed them into incompetence. The way to keep designers sharp and honest is to make them eat their own dog food. Any process that deprives designers of that feedback is ultimately doomed to failure."

from API: Design Matters

Micro-Services: The Modern Modular Programming Paradigm

Modular programming is a software design technique that emphasizes separating the functionality of a program into independent, interchangeable modules, such that each contains everything necessary to execute only one aspect of the desired functionality.

As Service Oriented Architecture (SOA) has become popular the modular programming definition could easily be rewritten to read:

The micro-services style of service oriented architecture is a software design technique that emphasizes separating the functionality of a service into  independent, interchangeable services, such that each contains everything necessary to execute only one aspect of the desired functionality.

The term modular programming was coined in the late 1960s. It has been a key part of good software engineering since.

Here are some positive points about micro-services style of SOA:

• Small, easy to understand code base. Because micro-service application is responsible only for one thing, it requires less code, is easy to understand, is easy to reason about, is easy to test, and has less risk of changes.
• Easy to scale. With a large, monolithic application you have to scale everything together. You have to scale the whole thing and waste resources. Whereas, if demand for a particular service temporarily increased then you would only need to scale the specific service.
• Easy to Deploy. With monolithic applications even one line code change require redeployment and re-testing of the whole application. This could be a big problem for many organizations with its high degree of risk and disruption. Deploying micro-services is much simpler because the scope of deployment is much smaller. Plus you know where to look for problems if such arise.
• Ability to use a different technology stack. With micro-services the approach should be to use the best tools and languages for the job, instead of one size fit’s all. The same goes for databases. For example, recommendation micro-services can use neo4j and python because python has a lot of machine-learning libraries; event-processing micro-services may use java and cassandra because of multithreading properties of jvm and high scalability of cassandra. It also allow teams to try new technologies on small services without major disruption.
  Smaller teams. It’s much easier and faster to work with a collocated, small team. Each small team can own micro-service and access other services via high level api. For example, a team in New York can be responsible for recommendation services and a team in India for content micro-service.
• System resilience. If an monoliths application stop working, then a lot of functionality stop working. On the opposite side, If one of the micro-services stop working – only small, particular functionality will be lost. It’s much more simple to build some resilience around smaller service. For example, failure of the order-processing system can be mitigated with message queues.

Regression Testing - Testing Variability

When you need to test that a code change has made an improvement or, at least, didn't reduce the current level of performance, it is called regression testing. A level of statistical analysis is needed to ensure that perceived differences are not the result of random chance. The rigorous way to accomplish this is to use the Student's t-test to compare the results. The t-test can tell you the probability that a regression exists, but it doesn't tell you which regressions should be ignored and which must be pursued. However, provided the cost of running additional tests a cost-benefit analysis could tell you that.

Check out the Apache Commons Mathematics Library (TTest) for an implementation for Student's t-test.

Get To It

"Our great business in life is not to see what lies dimly at a distance, but to do what lies clearly at hand." - Thomas Carlyle

My sister, Joyce, and I write letters back and forth. It is a bit of a tradition that we include a quote either in the letter or on the back of the envelope. The above quote is what I included with my most recent letter. I picked it because all to often at work it seems that we spend so much time worrying about what may come along in the future that we don't do what lies clearly at hand.


If that seems too short sighted to anyone then I implore you to remember the popular quote of Lao-tzu. "A journey of a thousand miles begins with a single step."

Strive To Fix What You Can

I will admit to sometimes getting frustrated at work when someone doesn't want to work in a specific codebase because they think it has too many problems. George Bernard Shaw said, "If there were nothing wrong in the world, there wouldn't be anything for us to do." The systems that need the most work don't just validate the need for developers but are an opportunity to make major improvements.

I had a friend in grad school who went to work fixing and upgrading Cobol applications at a large financial company. Much of the software had no major maintenance in years. He loved it. He said that everywhere he looked he found things to improve that made him look like a rock star.

When you find yourself in that legacy application that is out-of-date and has little to no documentation. Strive to make it better and people will notice. You might even change people's perceptions of the application.