DEV Community: Bilaniuc Dragos

Sketching out the extremes: an approach to designing software architectures in highly unpredictable projects

Bilaniuc Dragos — Mon, 12 Aug 2024 07:29:23 +0000

Imagine a scenario: a fresh, ambitious project with enormous potential but limited resources. The stakes are high, and the product needs to hit the market as soon as possible to validate its potential and seize the opportunity. The roadmap is hazy, making flexibility crucial.

Classic story until now. But here’s where things get spicy: the company that started the project has a huge community ready to jump on this new product they’ve been advertising for months. Thus, at the initial beta release, the system must be prepared to jump from 10 internal people testing it to anywhere from 20-100k daily active users and millions of events per hour.

This mix of conditions creates a particular set of requirements:

the system must be simple enough to be handled by a few devs at the beginning, with short iteration time and cost-effectiveness being key
it must be scalable enough to handle the spike of users around the initial release date
it must be flexible enough to permit a rapid move toward scalability later on, delivering more and more complex features along the way

The process

The first step was to create an architecture that would satisfy all these requirements.

As a preparation, I read and re-read the following books:

Drawing on the insights from these resources, along with my prior experience, I knew that some form of distributed architecture would be essential to handle the expected traffic.

Roughly speaking, I followed these steps to arrive at an initial sketch of how the system would look like:

Analyse requirements and gather events that will need to be handled by the system
Group those events into domain boundaries
Identify architecture characteristics (scalability, security, robustness - these kinds of things)

That initial sketch looked like this (in order to protect the intellectual property of the company, I anonymized all event and module names, keeping only the shape of the sketches):

Readable version here

After lots of refinement and discussion sessions, two things became clear:

one big chunk of code (i.e a monolithic architecture) wouldn’t fit our scalability and flexibility needs
a full-blown microservice architecture would be too complex and costly at this stage of the game

So, as with most things in life, the solution was somewhere in the middle. Finding that middle was now the challenge.

Here’s where the approach that I’m proposing comes into play:

Sketching out the extremes

Initially, the mindset was to design the simplest architecture that would satisfy our requirements and see if I noticed any specific areas of risk/improvement.

The simplest architecture I could think of was a modular monolith, where those modules would be as separated and clearly defined as possible. It looked something like this:

Readable version here

The thing that I thought would be the “secret sauce” here was the data access library, a library shared between all modules, it would sit in front of the database layer and would be designed in such a way that extracting a module into its own microservice would take as little effort as possible, and we would only have to tamper with the shared library.

As a bonus, and to be prepared “even more” for the move towards microservices, I was planning to deploy each module into its own independent function on the cloud infrastructure we were planning to use.

After analyzing the approach a little more, two things became obvious:

implementing the “secret sauce” library in the way that I imagined wasn’t realistic, and it would lead to lots of complications
deploying modules in different containers that accessed each other's databases risked creating a "Distributed Monolith"—a problematic architecture that combines the worst aspects of monolithic and distributed systems.

So, since I went to one extreme, i.e. the cheapest/fastest one, I thought it would be interesting to go to the other one as well:

What would the architecture look like, if we had all the resources we needed to create it?

After some more sketching, I arrived at this rough design of a microservice architecture for our system (I only highlighted some the inter-service events here):

Readable version here

In this version, we have some clear advantages:

individual scalability of each module, based on loads
clear ownership of the data
each module can be independently deployed

However, these advantages come at a cost: complexity.

Most of that complexity came from the inter-service communication. In V1 of the architecture, if module C needs data that is stored in database E, it can directly access it via the shared library. In this version, it has to request the data from Module C over the network - which adds more points of failure and makes tracking and debugging more difficult.

But even though this wasn’t the best solution for our current situation, it led us to some important insights based on which we built the actual architecture of the system.

The solution

After taking a closer look, we noticed that most of the inter-service communication was happening between 3 of the 6 modules: Module C, Module D, and Module F. If the network communication between those services was somehow taken out of the picture, the remaining microservices would have a reasonable number of events and requests to share between them.

So that is exactly what we did: compressed the 3 highly interrelated modules into a single modular microservice, and transformed the other modules into their own microservices.

After some more tweaking and refining, this is the architecture we’ve come to:

Readable version here

Here are the advantages of this architecture:

it retained the scalability advantages of microservices
the downsides of inter-service communication were reduced to a manageable level, enough so that our small initial team could implement the architecture in an efficient manner
the "big" microservice was implemented modularly, allowing for future extraction of individual modules if needed
the established microservice framework meant there would be less friction when splitting additional components into new microservices

Conclusion

Oftentimes, as software developers and architects, we are expected to come up with the ideal solution for any set of circumstances, requirements, or problems. We are aware, however, that everything in software development is a tradeoff.

While this little story isn’t about promoting any specific type of software architecture, its purpose is to showcase a promising strategy for building software architectures, and it sounds like this:

When trying to figure out where exactly to stop between two extremes, a good starting point is to sketch out how both of these extremes would look like.

In terms of software architecture, designing the cheapest and fastest architecture that would sort of satisfy the requirements, and putting it next to the most comprehensive and scalable solution can be very insightful. It can bring a better understanding of the system, its risks, and possibilities - as well as provide a clearer general direction for the product as a whole.

In our case, this strategy showed us how a little tweak in the way we were organizing our components allowed us to aim much closer to the “ideal” microservice architecture than we had initially thought was possible.

What insights would such an approach bring to your project and team?

How I stopped freaking out when speaking to people | A journey with 4 lessons

Bilaniuc Dragos — Fri, 23 Jul 2021 16:33:27 +0000

Imagine: You're in 6th grade. Your English teacher gives you the main role in a poetry event that will appear on TV. On the big day, you walk on the stage. The camera turns upon you. Everyone's eyes turn upon you. You take a deep breath, and… blank.

I always like to point to that story as the reason for my excessive fear of speaking in front of people. Of course, there must be lots of other reasons for it, but having a brain fart at 11 years old - on TV and in front of everyone - sounds dramatic enough, so we'll stick to that. The fact is, this became one of my deepest fears.

Learning to live with this fear - and (spoiler alert) going from that panicked little kid to winning the Speaker of the Year award in front of over 300 people - was the most transformational journey of my life. It taught me some of the most important lessons I know, so I decided to write about them. Maybe they'll help you, too :)

Denial

For as long as I can remember, whenever I'd have to speak in front of people, my heart would start thumping, my mind would start racing and I'd start blurting out words rapidly and with no coherence. And the fact that I was in complete denial of this fear didn't help at all.

I quickly fell into a very nasty loop: I'd try to mask my insecurities with "confidence hacks" from youtube, people around me would smell that and become even more repulsed, which confirmed my insecurities, which in turn made me seek even more tricks and hacks… and so on.

And this loop led me into a really dark place. I couldn't feel an authentic connection with anyone - not because there was anything wrong with the people around me, but because I never allowed them to see who I really was.

All I was doing was keeping up this facade of fake confidence, and getting disappointed and frustrated every time it would inevitably fail to solve my problems.

Change

The trigger for change came at a leadership event that I participated in. I was my usual self, feeling awkward and trying to mask it with tricks and hacks.

And then, there was this guy - speaking calmly, telling the right jokes at the right times, never really "trying" anything, but making everyone feel good - especially when he got in front of the group. Everyone wanted to be around him. And I wanted to be him.

Following his advice, I joined Toastmasters, an organization that had regular meetings for practicing public speaking - and this is where the real pain of confronting my fears and insecurities began.

At first, I was obsessively practicing my speeches (30–40–50 times!), just to make sure what happened in 6th grade would never happen again. And it went quite well, insofar as blocking on the stage went.

But the painful part came after a speech was over, and I would look over the recording. I could barely stand to look for more than 2 minutes - that's how unnatural and awkward I saw myself on the stage. All my insecurities were exposed in front of everyone. And that was a hard pill to swallow.

The only thing that kept me going was gradually learning the lessons I'll talk about. The made me stick to this journey, one painful step at a time, until I was holding this speech:

that ended up with me holding this beauty, as one of the biggest achievements of my life:

Now, on to the lessons…

1. The most powerful thing to know is that it's okay to fuck up

And I'm not saying it in a make-me-feel-better way. This is actually the best way to minimize the mental chatter that's such a big part of fear.

Most times, fear would begin with the question "what happens if I fuck up?", which would trigger an endless loop of apocalyptic scenarios in my head - scenarios that were totally ungrounded in reality. Once I accepted that it's okay to fuck up, I could short-circuit the loop by answering that question with a realistic "nothing much".

Toastmasters taught me that environment can play a crucial role here. If the feedback for my first speech was an honest "you suck, just go home" - you can bet I'd still be filling the pockets of "hack your confidence" gurus on youtube.

Instead, the message was always: it's not a big deal - it happens to everyone. And that's about the most empowering message you can get when confronting fear and insecurities.

Another thing that really helped me with this point was seeing other people around me awkwardly trying to confront the same fear and fucking up themselves. Knowing that I was not the only one struggling (and failing) was what kept me going, during some of my most intense moments of self-criticism.

And this brings me to my second eye-opening lesson:

2. Encouraging others is the best way of dealing with your own fear

And this is something I didn't expect at all: the most effective method of dealing with my own fear didn't have much to do with me. It was all about encouraging others to confront their own version of this fear of public speaking.

And not just because it's the right thing to do in a higher, humanitarian way. It's actually the best thing you can do, egotistically speaking. Because when you see someone else dealing with the same kind of fear you're dealing with - and start encouraging them - something amazing happens:

that fear becomes smaller for both of you

This was another huge advantage at Toastmasters: being a member, I didn't only give speeches. I would also have to evaluate others, encourage them and give them the best feedback I was capable of. And telling others that they can do it, I suddenly found myself believing it, too!

Find someone that's dealing with the same kind of fear. Encourage them and let them know that it's okay to struggle and even to fail. Then feel the empowerment this simple act brings upon you, in your own struggles.

And this lesson kept me fighting, until I finally realised that:

3. You don't conquer fear. You learn to live with it

About 2 years passed since I won that trophy that marked the success of my "public speaking project". Since then, I've had quite a lot of events to moderate and speeches to present. In short, I kept gaining experience.

But just the other day, I was going to this salsa&bachata party where I didn't know anyone. I arrived a little earlier, and there were a few people gathered up in a discussion. As I approached them, my heart started thumping, my thoughts started racing - all the good old friends were right back.

And this is the biggest takeaway from my journey:

My fear never went away. I just learned to live with it.

More specifically, I learned to expect fear, without expecting that it will ruin everything. To leave it in the background, and keep going anyways.

I believe I learned to do this simply through experience. Speech after speech - no matter how much I rehearsed or how small the audience was - the same feelings of fear would arise. And, speech after speech, they would prove to be less tragic than I imagined.

Eventually, those feelings became just part of the landscape - not going away, but not consuming all my attention, either. And that's how I started to actually enjoying public speaking. That's how my gruesome fight came to an end- not because I won or lost, but because I stopped fighting.

What I wish I knew then is how much mediation can help with this - since its core training is leaving thoughts and emotions in the background. And it's such an effective way of doing it that, in hindsight, I believe it would've saved me months of pain and frustration I had to go through, just to begin to learn this lesson.

Just for the record, after trying out lots of apps and techniques, the most effective way of meditating for me right now is through the guided sessions of Jeff Warren, available on Calm. If you have a hyperactive mind like mine, he might resonate with you, as well :)

And the last big lesson this journey had to teach me was that:

4. It's absolutely worth it

I believe that one of the most perverse things when it comes to dealing with fear and insecurities is how easily the battle can be lost before it even begins.

For such a long time, all I did was find circumstances and excuses. I'd tell myself it's alright, I'll eventually cover those insecurities up with confidence hacks. Or I'd have a little fuckup and see it as confirmation that I'll never be able to do it. Or, most often, I'd just delay things, expecting some enlightenment moment where all my fears and insecurities would disappear, just like magic.

And now, I can confidently say that these were the worst ways in which I was sabotaging myself. Because learning to live with only this one fear, changed the way I looked at every other area of my life.

More and more often, when faced with insecurities, fears and uncomfortable situations, I found myself thinking:

"if public speaking didn't kill me, this won't kill me either"

And soon enough, this automatic way of thinking made me take on opportunities that would've terrified me before - opportunities that, among others, ended up sending my professional growth through the roof.

I gained the courage to begin freelancing, to assume the role of Tech Lead in my projects, and, soon after, to become the CTO of a very promising startup.

All of these moves came with lots of doubts, fears and insecurities packed in. And the only way I could get over them was through the lessons learned in this seemingly unrelated journey of confronting my deep fear of public speaking.

Final thoughts

So, that's about it for my journey. Confronting my fear taught me some of the most valuable lessons I know: that it's okay to fuck up, that encouraging others pays huge dividends, that you can live with fear and that it's totally worth doing it. I hope you found some value in those lessons, as well.

But, at the end of the day, this is just one journey among millions. Fear and insecurities are everywhere, and each of us is trying our best to deal with them. If you're struggling, rest assured that you are not alone, not by a long shot.

And if my specific lessons don't work for you, what will definitely work is what you're already doing now: keeping an open mind, seeking out new ideas and enlarging your own perspectives. A little bit from here, a little bit from there, and that's how you're building up your own recipe for dealing with this existential human condition.

Kudos to you, for keeping an open mind in front of so many fears and insecurities!

I tracked every minute of my time for the last 4 months. Here are 7 totally unexpected results

Bilaniuc Dragos — Wed, 14 Jul 2021 13:11:09 +0000

A 30 minutes lunch break? Time entry. 12 minutes of arguing with mom? Time entry. 7 minutes of stalking my ex on Instagram? You guessed it!

That was my life over the last 4 months, being my own time tracking cop while carrying out the most outrageous experiment of my life so far. The results amazed me in ways that I would never have expected, so I decided to write about them. Who knows, maybe you’re also going to find something of value in my journey.

But first…

What the hell made me do it?

To be honest, it was a combination of frustration, guilt and fear — all generated by the fact that I kept setting big goals for myself, but never seemed to make any progress towards them. Time was passing, and I had nothing to show for it.

All I had were vague reports at work, where “working” covered everything from writing code, to talking to people, to watching Russian slapping contests on youtube.

I knew I had to do start tracking my time if I wanted to keep my sanity. And I’ve seen how something casual/half-hearted would quickly become too vague to be of any use. If I wanted to do it, I had to go all-in.

So…

How exactly did I do it?

I used an app called Toggl to do all my tracking. I chose this specific app because:

it’s available on all platforms and syncs very nicely between them
it has an intuitive UI, with suggestions that become more and more accurate over time
it allows for very fine customization and reporting

At first, I devised 7 main areas where all my entries would fall, using tags and colors to mark them in the app:

Then, I created some high-level projects that I knew I was already spending time on:

After the setup was done, I would simply start a new time entry, every time I would move from one activity to the other.

One important note here was that, because I decided to go all-in, I had to be particularly careful not to have any entries “cover up” unrelated things that I was doing. For example, if I would be writing code on my project and I’d receive a call, I’d have to create a separate time entry for it, even if it only took 5 minutes.

This level of strictness was a bit taxing at the beginning, but after a while, Toggl’s UI came in very handy. I rarely had to create new entries. Most of the times when I’d switch to something else (even just short interruptions), I would just have to press “continue” on a previous entry.

At the end of the week, here’s how my timetable would look like:

And now that the “how” is also out of the way, let’s move on to the exciting stuff. Here are the biggest changes I noticed, over these 4 months:

1. Drastically improved my focus and productivity

Yep, the biggest impact of this time tracking journey didn’t come in the form of a mind-blowing insight at the end of the week, when I was looking over my timetable. The biggest impact was in how I was spending my time while I was doing the tracking.

At first, nothing big happened. But after a while, I found myself more and more often in an interesting situation. Whenever I was working and felt the need to “zone out” by jumping “just a bit” on youtube or Instagram, I’d find myself thinking “nah, I don’t want to start a new entry now. I’ll just get this thing done, and then take a bigger break for that”.

And, more often than not, I wouldn’t even feel that need for youtube or Instagram anymore, once I actually finished the task. The feeling of accomplishment was enough to give me the refreshment I needed.

This increase in focus sent my productivity through the roof. Yes, I’ve always read and heard of how context switching is a very bad idea. But, until I tracked it down, I never knew how often I was doing it — and I never knew that getting rid of it would make me get things done literally twice as fast! After seeing all this for myself, I felt like I somehow hacked my own life.

2. Saved a lot of time from unexpected places

One of the most shocking insights that I had in the beginning, after looking at my reports for a week, was that I was spending over 20 hours a week on food-related activities (buying food, cooking, eating).

Every day, 3 times a day, I would take some time off to cook. And, since eating healthy was always an important thing for me, I didn’t feel bad if it sometimes took a little longer. But I never knew I was doing the work of a part-time cook out there! At least if I’d learned to cook something more exciting than this boring salad:

Once I became aware of this, I automatically started thinking about how I could make it a bit more efficient. A simple thing that popped into mind was that, instead of cooking everything from scratch in each of the 3 breaks, I could simply prepare everything in the evening before, and have my food ready to go for the next day, with minimum adjustments.

This trick solved all my cooking for the day in under 1 hour, instead of 2–3, which amounted to over 10 hours of saved time, per week. And the best part was that I didn’t have to sacrifice anything in terms of how healthy or tasty my food was!

Such a simple solution. However, I would never have thought of it, if there wasn’t this table that was telling me straight to my face: “here’s how much time you’re actually spending on this thing”.

3. Formed habits & routines much easier

Turns out time tracking helped with my habits and routines, as well, in two not-so-obvious ways:

1 - If I would have the same entries at roughly the same hours, Toggl’s mobile app would start suggesting them, in those time intervals. This made my tracking way easier, and turned the clock into a kind of trigger for the habits I was trying to form:

2 - There was a deep satisfaction when I would look at my weekly timetable and see all the entries nicely aligned. Here’s one of the weeks that I’m most proud of. Just look at how satisfying those morning entries look:

It might sound trivial, but just these two simple tricks were enough to make me form and cement very helpful habits, without thinking too much about them.

4. Enjoyed my “chill time” more than ever

Here’s something interesting. This experiment, which was chiefly meant to increase my productivity, had another very unexpected — but very welcomed — effect: it increased the quality of my “chill time”, as well!

Firstly, as I was doing my work much more efficiently and saving time from all kinds of unexpected places, I found myself with more and bigger blocks of time on my hands. That meant I could fit in some leisure activities that were more time-consuming — but that I really enjoyed doing.

Secondly, all this increased amount of “chill time” came without the usual guilt I felt before, when I was “chilling” by scrolling through the Instagram feed, while I knew I had to fix that stupid programming bug at work.

This made me enjoy my free time like never before, and rekindle old passions like salsa & bachata —passions that I forgot how good they made me feel, amongst all this clutter of work and guilt and continuous distractions.

5. Made better decisions, overall

one of the toughest decisions in life

Since I was being as honest as I could with those time entries, weighing my decisions wasn’t that complicated anymore: I would look over my timetable, see exactly how much time I spent on something I decided to do, and look at how much I got out of it.

This way, “Is it a good idea to write that article for my 23rd birthday?” became “Is it a good idea to write that article for my 23rd birthday, given the fact that last week I spent 11 hours, and got nowhere with it?”. Not so hard to see the issue, from that perspective.

As a bonus, tracking my time solved another one of the biggest problems I had, in terms of decision making. I always had the tendency to overestimate my abilities and take on much more than I could actually handle. This obviously had some very bad consequences, like frustrating other people that were working with me.

Well, once I had this sheet that was telling me “this is how much it actually takes you to do it” and “this is how much free time you have on your hands right now”, deciding whether I could or could not handle something new became a factual decision, rather than a whim of my arrogant ego.

6. I finally felt in control of my time

As I said in the beginning, the irreversible passing of time was one of the biggest sources of stress and anxiety for me. And, while I’m still having my fair share of existential crises about it, doing this experiment drastically reduced that stress and anxiety. The main reason was that, again,

I finally felt in control of my time.

Yes, I was still binge-watching cat videos on youtube, probably more than I should. But at least it was a conscious decision, rather than a reaction to some algorithms designed by smart people to make me mindlessly spend hours on end on their website. And this “conscious decision” part was what made all the difference, for me.

And the coolest thing was that I didn’t have to change anything in the way I was spending my time, in order to feel better about it. Simply having it noted down was enough to give me that empowering feeling of control.

7. Felt more tense, uneasy and guilty (at first)

Surprise-surprise: this is not a silver bullet to shoot down all of life’s problems. It actually came with some not-so-nice consequences that I have to mention, if I am to paint an honest picture of my journey.

First up, I became more tensed and uneasy, especially in the beginning. I think that’s not too unexpected — just imagine having a cop on your back, 24/7, that sees every single thing your doing, and tracks it down on a sheet that’s gonna stare you in the face every Sunday. Quite a reason to feel uneasy…

shoutout to Mr. policeman, for not arresting me when I randomly asked him to pose for my blog

The other thing was guilt. Again, especially in the beginning. After the first few weeks, I was confronted with a hard truth: I was wasting time. Lots of it. And not in a vague way. I could actually count the hours I spent on things that were totally unproductive, projects that failed and times when I did something completely different than what I had to. Talk about pills that are hard to swallow…

I eventually managed to keep those bad sides under control through other habits, like meditation. That being said, this journey was definitely not all fun and rainbows.

So, that’s about it for my journey. Currently, I got so used to this system and it seems to work so well for me, that I’m gonna keep doing it — as long as I’ll be able to keep those negative sides under control.

Afterthought: it’s not as hard as it seems

I’ve been showing this article to some friends, to get some feedback before publishing it. A common response was: “Wow, that’s so impressive! You must be so disciplined! I could never do that.” And while it tickled my ego to hear this, the truth is, it’s really not that hard.

Yes, like all things, it requires effort in the beginning. But after you got it all set up and run it for a few days, it becomes pretty easy. 99% of the time, you’ll just press “continue” on a previous entry, which only takes 5 seconds. And since you’ll have to do it dozens of times a day, it will become instinctive much sooner than the usual “3–8 weeks to form a new habit”.

Even now, my timer runs “Blog — time tracking” for 9 minutes and 20 seconds. I don’t remember having made a conscious decision to start it.

Tips & tricks

If you like the idea and consider trying it for yourself, here are some things I wish I knew when I started, and that I hope will help you in your journey:

Try to be as honest as possible. The benefits are gone if you find yourself scrolling through your Instagram feed, while your timer says “preparing yearly report”
Keep it personal. Don’t mix it up with any tracking software from work. Don’t tell anyone about it. Basically, don’t do anything that would make you feel even remotely uncomfortable about recording those 2 hours of binge-watching Japanese eating contests on youtube.

Keep it simple. As simple as possible. Don’t overkill it with tags and projects (the way I did, in the beginning). You can always add those later. The whole system should be a pleasure to use, or you’ll have trouble using it at all.
One step at a time. At first, don’t change anything in how you spend your time. Just get used to tracking down what you’re already doing. The main focus is to get used to pressing “play” every time you’re doing something new.
Be patient with yourself. Once you start doing this in full honesty, you’ll probably see that your time isn’t really spent the way you imagined. That’s an easy way for guilt to creep in. But don’t despair. Once you become aware of this, things will get better by themselves. And keep in mind that, just by trying it out, you’ve already done something very courageous that’s a huge achievement in and of itself.

Conclusion

Tracking every minute of my time is, by far, the most effective thing I’ve ever tried, in terms of becoming more focused, getting more done and improving the quality of my time.

But, at the end of the day, this is just another way in which someone on this planet is trying to deal with the fact that life is hard. And if my trick doesn’t work for you, what will definitely work is keeping this healthy mindset that makes you read these kinds of articles, in the first place. Kudos to you, for keeping an open mind in the face of life’s hardships!

We’re all trying to make the most out of our time. Thank you for spending some of yours with me, today!

PS: This is the first article I’ve ever written, so I think you can imagine how anxious I am about it 😬 Any though (positive or constructive), clap, comment, share, or feedback in any form would be very, very much appreciated!