Celestial Bodies and Micro-Front End Reuse

Source: Wikimedia Commons

If you are writing a modern production system of any complexity, chances are you are using microservice architecture for it. If that system serves the Web browser, sooner or later you will discover that you must break up the UI into parts served by multiple micro-services (or micro front-ends).

The promise of microservices is that they can be built by separate teams, allowing teams relative independence and ability to move at their own speed, without the need to for intergalactic coordination. However, sooner or later executives (God bless them) will notice that some of the user interfaces contain elements that look similar across the system, and the dreaded word “Reuse” will raise its head.

It can sound downright heretical to say anything against reuse in software industry. That code reuse saves time and improves quality feels so innately right that it takes gumption to raise against it. What would be more logical than to create a reusable component for things that look similar and make all micro front-ends reuse them?

And yet.

Movement on a planetary scale

To try to illustrate the problem we can hit with indiscriminate reuse, let’s look at the movement of planets in a solar system. As planets rotate around the star, they follow their own orbits that place them at various distances to each other. If we simplify the geometry and make all orbits perfect circles and in the same plane, the planets will still have different rotational speed, dictated by various factors (yes, ‘Gravity’ violated many of them). This rotation will place them at various positions to each other, with two extremes:

  1. Conjunction – when two planets are lined up on the same side relative to the star and closest to each other
  2. Opposition – when two planets are lined up on opposite sides relative to the star

As planets are in constant motion, they will at any moment be between these two points.

Back to the topic

OK, so let’s bring it back to the microservice system. The reason UI component reuse is being discussed in your meetings is because a design pass was made, and the first iteration of the whole system’s user experience has been put together, and some repetition is observed. This page in micro front-end A looks very similar to the page in micro front-end B. That picker is the same in three pages. Let’s solve it once and for all for everybody. Reuse, people! Sensible leadership all around.

It is easy to fall into this trap because reuse works great for, say, libraries. Of course you will not write that set of mathematical transformations over and over everywhere. In our Node.js microservice system, the list of modules each app depends on is getting longer and longer.

I just typed npm list > deps in one of our largest microservices (I am tempted to call it Gustav for its girth). The flattened list of all dependencies is 3689 lines long. We should call it “miniservice” or just “service” by now. I am not writing 3000 modules by hand any time soon. Reuse is fine and dandy here.

Similarly, all our micro front-ends are using the same platform API layer. The APIs are ensuring the state of the microservice system is consistent. Again, I am not saving the same state in two APIs and keeping it in sync somehow – that would be crazy.

All our micro front-ends are using the same style guide and the same set of React components written to follow that guide with a set of basic visual building blocks (no, we are not using Bootstrap; yes, we know 15% of all web sites use it – good for them). Without it, our UI will look like a ransom note:


Say NO to casual UI reuse

This brings us to reusing actual parts of pages between micro front-ends. If you follow the reasoning above, it is the logical next step. This is where our astrophysics comes into play.

The pages that look similar today are like planets in conjunction. They are at their closest position in the microservice planetary system. The danger is that we are forgetting the constant movement. It may be hard to believe right now, but these pages will move apart. Why? Complex systems rarely evolve in perfect coordination. You cut your hair and a month later the cut starts losing shape, because the hair refuses to grow exactly the same everywhere. Diverging priorities, architectural changes, business pressures, reorgs all conspire against your microservices keeping their relative distance to each other.

The whole point of a microservice system is to insulate against this corporate chaos by allowing self-sufficient teams to cut through the haze and continue to make progress. Wonton UI reuse will tie the microservices together and restrict their freedom of movement. This is an elaborate way to write a distributed monolith, the worst kind. In our planetary analogy, you are applying super-natural gravitational forces between planets that completely mess up their trajectories and may result in the fiery collapse of the entire system. Or ‘Gravity 2’.

Doing it right

The heretical thought presented in this article is that some duplication and bespoke UI code is good if the outcome is preserved agility and independence of the microservice teams. Of course, thousands of lines of duplicated code is its own kind of a waste, agility or not. This is where good practices are essential, such as using back end for front end to push most of business logic to the server, leaving mostly presentation logic on the client. The business logic can then reuse libraries and modules across microservices (within reason, you need avoid the danger of increased coupling so don’t go overboard).

If you do it right, you will already use common style, common APIs, shared libraries and common widgets. The rest is just composition, where you can create a bespoke UI that perfectly fits your page and nobody will break you by pushing a change without enough testing at 2am.

Any exceptions?

Yes, there are exceptions. Some UI components are just too tricky to implement even once, let alone multiple times. You will know those when you see them – the microservice teams will come together and agree that one of them will build it for all of the teams because life is too short to fix that horrible bug 5 times.

The threshold to clear will be:

  1. Is this truly a single component, or multiple components merged together, selected with a maze of flow control and configuration parameters? If latter, don’t.
  2. Is the component simple enough and with a single purpose, more likely to survive inevitable zigs and zags of business requirements and design changes?
  3. Is the team writing the shared component up to the task and ready to write exhaustive functional and UI tests to guard against regressions and breakages?
  4. Is the problem complex enough to overcome the overhead of reuse?

Once you are true to these conditions, reusable components will be few and far between. If a once useful reused component becomes a sore point in a year, fork and take the control back.

You will reap the consequences

In complex microservice systems, some of the truisms of software development need to be put to the test and re-examined. The true test is not if applying them works at the current moment (at the planetary conjunction), but if the system will survive the opposing forces of changing business requirements, architecture evolution, aging code and organizational changes.

As engineers, we must remember that business defines requirements, design defines the UX, but we are responsible for turning the requirements and the design into clickable reality. It is our responsibility to choose that reality not only based on what our system looks like today, but what it may look like in 6, 12 or 18 months. Resist the urge to reach for quick shortcuts through casual reuse, and plan for what situation that reuse will likely put you after a few architectural and organization changes.

Like planets, your microservices will never stand still.

© Dejan Glozic, 2017






Microservice Grid and Micro Frontends

Adam naming the animals, Wellcome Library, London. Source: Wikimedia Commons.

This post is not here to bring you anything new as it is for me to run a victory lap. If there is anything I blogged about with any consistency, it was about microservices. There are many articles in this blog addressing various problems you may face when attempting a microservice system of a non-trivial scale. One area I was particularly keen on was decomposition of the front end in a way similar to what we did in the back.

You would think that it goes without saying, but a lot of microservice noise was focused on the backend. I understand that: breaking the big server into a number of smaller ones behind a routing proxy is easier to understand. REST API is REST API is REST API. A number of endpoints per microservice neatly collate into a complete system with use of session-less API services.

But what’s good for the goose ought to be good for the gender. Slapping a massive, unruly front end server in front of a nice microservice API system does not help us much, other than to highlight the irony of the half-hearted approach. That’s why I focused a lot of this blog on composing the UI microservices as well. From your favorite article on microservice authentication, to UI composition, to isomorphic microservices using Node.js and React.

Far from being armchair generals, we put our money where our mouth is, applying these ideas in production, learning more in each successive generation. The last name for this approach I used was microservice grid:

microservice-gridThe picture above represents the essence of our current approach. It is uniform in that horizontally it represents tiers (front end, back end, optional hosted legacy apps in containers). Three buses provide glue – UI domain for bringing all paths under the same URL, API domain for collating all API endpoints, and an event bus to glue the microservices using a message broker of some kind.

Vertically, each column of the grid represents a feature. What I like about this approach is that unlike with big monolith apps, adding features does not make everything bigger and slower. Adding a new feature vertical does not affect the performance of the existing verticals.

So that’s what you call it

One of my favorite hobbies now is going to Technology Radar published by ThoughtWorks and looking for how they are naming a technique we are already using in production. For example, the pattern of providing a single API gateway for your client side code is now called Back End for Front end. But what gave me no end of pleassure was that now the whole approach of this blog that I spent so much time writing about has an official name: micro frontends.

We’ve seen significant benefit from introducing microservice architectures, which have allowed teams to scale delivery of independently deployed and maintained services. However, teams have often struggled to avoid the creation of front-end monoliths—large and sprawling browser applications that are as difficult to maintain and evolve as the monolithic server-side applications we’ve abandoned. We’re seeing an approach emerge that our teams call micro frontends. In this approach, a web application is broken up by its pages and features, with each feature being owned end-to-end by a single team. Multiple techniques exist to bring the application features—some old and some new—together as a cohesive user experience, but the goal remains to allow each feature to be developed, tested and deployed independently from others. The BFF – backend for frontends approach works well here, with each team developing a BFF to support its set of application features.

Technology Radar, Nov 2016, ThoughtWorks

Like with Adam in the engraving above, things exist before they are named, but Nomen est Omen – naming our approach in a way that offers a nice symmetry with micro services gives it a renewed meaning and is a success in my book.

Therefore, go forth and build your microservice grids, your micro frontends, your Backends for Frontends, you princes of Maine, you kings of New England.

© Dejan Glozic, 2017

The Web of Hooks


It’s a new year, which means it is time to update the copyright at the bottom. I have been busy lately and I didn’t want to write fluffy blogs just to fill the space. But this time I actually have real content, so here it goes.

If you followed my blog in the past, you know I was very bullish on message brokers in the context of microservice architecture. I claimed that REST APIs alone are not sufficient to sustain a successful microservice system. Event collaboration pattern is necessary to ensure a scalable and robust system where microservices that own resource lifecycle don’t need to be burdened with executing all the code driven by that lifecycle. You can read more about it in my article about REST/MQ mirroring.

Not so fast

There is a fly in this particular ointment, however. While HTTP REST is well understood, messaging protocols are numerous. You may know from my writing that I liked MQTT for a while due to its simplicity and wide support. However, MQTT is actually not a great protocol for gluing your microservice system. It is designed to be lightweight and run on the smallest of IoT devices, and lacks some critical features such as anycast support and manual acknowledgement of messages.

Another popular protocol (AMQP) suffered a schism of sorts. The version supported in a popular RabbitMQ broker (0.91) is a very different protocol from an actual open standard AMQP 1.0 that is not as widely supported. This is a pity because I really like AMQP 1.0. It is like MQTT but with anycast and manual acknowledgment – perfect as a microservice system glue, yet still very simple to work from within clients.

And of course, there is Apache Kafka, a very powerful but odd choice given its origins in high-throughput distributed log aggregation. Kafka is unapologetically Java-centric, has a proprietary protocol and bolt-on REST APIs to connect client languages are still fairly low level. For example, where AMQP 1.0 guarantees quality of service and requires implementations to provide a buffer of messages that are re-delivered to a client that crashed, Kafka simply allows you to maintain a pointer in the queue but it is your job to work up the queue and catch up after restarting. You pay for performance by working at a level fairly low for general purpose application messaging.

Authentication pains

Choosing the messaging solution and protocol is only one part of the problem. If you want your system to be extensible and allow third-party integrations, you need to make it fairly easy for new clients to connect. On the other hand, you need to secure the clients because you don’t want a rogue app to eavesdrop on the events in the system without authorization (otherwise you have this egg on the bot face). Maintaining a large list of clients connected to the same topics can also be an scalability issue for some brokers.

For all these reasons, it has been widely acknowledged that message brokers are not a good match for external integrations. A cursory scan of popular cloud applications with large ecosystems all point at a more client-friendly alternative – WebHooks.

WebHooks to the rescue

You would think that if something is so popular and has a catchy name, there is an actual well written protocol for it. Wrong! WebHooks is the least common denominator you could imagine. In a nutshell, this is all there is:

  • You publish a list of valid events for which you will notify clients
  • You provide an API (and/or UI) for clients to register URLs for one or more of those events
  • When an event happens, you execute HTTP POST on the URLs registered for that event type.

That’s all. Granted, message brokers have topics you can publish and subscribe, and the actual messages they pass around are free-form, so this is not very different. But absent from WebHooks are things such as anycast, Quality of Service, manual acknowledgement etc.

I don’t intend to go into the details of what various implementations of WebHooks in apps such as GitHub, Slack etc. provide because thankfully Giuliano Iacobelli already wrote such an article. My interest here is to apply this knowledge to a microservice system we are building and try to anticipate pros and cons of going with WebHooks.

What it would take

First thing that comes to mind is that in order to support WebHooks, we would need to write a new WebHook service. Its role would be to accept registrations, and store URL and event type mappings for subsequent invocation. Right there, my first thought is about the difference between external and internal clients. External clients would most likely use the UI to register a URL of their integration. This is how you register your script in GitHub so that it runs on every commit, for example.

However, with internal clients we would have a funny problem: every time I restart a microservice instance, I would need to register somewhere in startup. That would make a POST endpoint a nonstarter, because I don’t want to keep creating new registrations. Instead, a PUT with a client ID would work better, where an existing registration for the same ID would just be updated if already there.

Other than that, the service would offer a POST for a new message into the provided event type that would be delivered to all registered URLs for that type. Obviously it would need to guard against 404s, 502s and URLs that take too long to return response, giving up on them after a set timeout.

The best of both worlds

The set timeout brings back the topic of the quality of service, implying that WebHooks are great for external integration but not that great for reliable glue of a microservice system. Why don’t we marry the two then? We could continue to use message broker for reliable delivery of internal messages, and hook it up to a WebHook service that would notify external integrations without the need to support our particular protocol, or get too much access into the sensitive innards. Hooking up a WebHook service to a message broker would have the added benefit of buffering the service itself so that it can be restarted and updated without interruption and missed events.


In the diagram above, our microservice system has the normal architecture with a common routing proxy providing a single domain entry into the microservices. The microservices use normal message broker clients to publish to topics. A subset of these topics deemed suitable for external integrations is also listened to by the WebHook service, and for each of those messages it reaches into the stored list of registrations and calls HTTP POST on the registered URLs. If the WebHook service crashes, a reputable message broker will maintain a buffer of messages to re-deliver them upon restart. For performance reasons, WebHook service can choose to keep a subset of registrations in the in-memory cache depending on how frequently they are used.


Obviously registering a URL with an HTTP PUT is much easier to implement, and providing a single POST endpoint to handle the event lowers the barrier of entry for external integrations. In fact, hooking up code to react to a single POST could very well be done using serverless architecture.

Are we losing something in the process? Inserting another service into the flow will add a bit of a delay but external notifications are normally for events that are not happening many times a second, so the tiny delay is more than acceptable tradeoff. In addition, if the client providing WebHook URL is itself load-balanced, this delivery will be hardcoded to anycast (the event notification will only hit one of the instances in the cluster).

Finally, this creates two classes of clients – ‘inner circle’ and external, segregated clients. Inner circle is hooked up directly to the message broker, while the external clients go through the service. In addition to this being an acceptable price to pay for easier integration, it is useful to be able to only expose a subset of events externally – some highly sensitive internal events may only be available to ‘trusted’ clients subscribing to message broker topics and having internal credentials.

Since the WebHook service will normally not keep retrying to deliver an event to an unresponsive URL, it is possible to miss an event. If this is a problem, external system would need to fashion a ‘belt and suspenders’ fortification, where event driven approach is augmented with a periodic REST API call to ‘compare notes’ and ensure the baseline it is working against is up to date.

© Dejan Glozic, 2017

Components Are (Still) Hard on the Web


Here’s Johnny! I know, I know. It’s been a while I posted something. In my defence, I was busy gathering real world experience – the kind that allows me to forget the imposter syndrome for a while. We have been busy shipping a complex microservice system, allowing me to test a number of my blog posts in real life. Most are holding up pretty well, thank you very much. However, one thing that continues to bother me is that writing reusable components is still maddeningly hard on the Web in 2016.

Framing the problem

First, let’s define the problem. In a Web app of a sufficient complexity, there will be a number of components that you would want to reuse. You already know we are a large Node.js shop – reusing modules via NPM is second nature to us. It would be so nice to be able to npm install a Web component and just put it in your app. In fact, we do exactly that with React components. Alas, it is much harder once you leave the React world.

First of all, let’s define what ‘reusing a Web component’ would mean:

  1. You can put a component written by somebody else somewhere on your page.
  2. The component will work well and will not do something nasty.
  3. The component will not look out of place in your particular design.
  4. You will know what is happening inside the component, and will be able to react to its lifecycle.

First pick a buffet

Component (or widget) reuse was for the longest time a staple of desktop UI development. You bought into the component model by using the particular widget toolkit. That is not a big deal on Windows or MacOs – you have no choice if you want to make native applications. The same applies to native mobile development. However, on the Web there is no single component model. Beyond components that are intrinsic part of HTML, in order to create custom components you need to buy into one of the popular Web frameworks first. You need to pick the proverbial buffet first before you can sample from it.

In 2016, the key battle is currently between abstracting out and embracing the Web Platform. You can abstract the platform (HTML, CSS, DOM) using JavaScript. This is the approach used by React (and my team by extension). Alternatively, you can embrace the platform and use HTML as your base (what Web Components, Polymer and perhaps Angular 2 propose). You cannot mix and match – you need to pick your approach first.

OK, I lied. You CAN mix and match, but it becomes awkward and heavy. React abstracts out HTML but if you use a custom component instead of the built-in HTML, React will work fine. All the React traits (diff-ing the two incremental iterations of virtual DOM, then applying the difference to the actual DOM) works for custom components as well. Therefore, it is fine to slip in a Web Component into a React app.

The opposite does not really work well – consuming a React component in Angular or Polymer is awkward and not worth it really. Not that the original direction is worth it necessarily – you need to load Web Components JavaScript AND React JavaScript.

Don’t eat the poison mushroom

One of the ways people loaded components in their pages in the past was by using good old iframes. Think what you will about them, but you could really lock the components down that way. If you load a component into your own DOM, you need to really trust it. Single origin policy and CORS are supposed to help you prevent a component leaking data from your page to the mother ship. Nevertheless, particularly when it comes to more complex components, it pays to know what they are doing, go through the source code etc. This is where open source really helps – don’t load a black box component into your DOM.

The shoes don’t match the belt

One of the most complex problems to deal with when consuming a Web component of any type is the design. When you are working in a native SDK, the look and feel of the component is defined by the underlying toolkit. All iOS components have the ‘right’ look and feel out of the box when you consume them. However, Web apps have their own themes that create combinatorial explosion. A reusable component needs to do one of the following things:

  1. Be configurable and themeable so that you can either set a few parameters to better blend it into your style guide, or provide entire template to really dial it in
  2. Be generic and inoffensive enough to be equidistant from any parent theme
  3. Be instantly recognizable (think youtube player) in a way that makes it OK that it has its own look and feel.

A very complex reusable component with a number of elements can be very hard to dial in visually by consumers. In corporations, this may reduce the number of themes you want to support. A large component may take it upon itself to support 2-3 supported and widely used design style guides. Then all you need to do is provide a single parameter (style guide name) to make the component use the right styles across the board.

What is going on inside?

Adding a component into your page is not only a matter of placement and visual style. Virtually all reusable components are interactive. A component can be self-contained (for example, all activity in a youtube player is confined to its bounding box), or expected to interact with the parent. If the component must interact with the parent, you need to consider the abstraction chain. Consider the simple countdown timer as a reusable component. Here is how the abstraction chain works:


The timer itself uses two low-level components – ‘Start’ and ‘Stop’ buttons. Inside the timer, the code will add click listeners for both buttons. The listeners will add semantic meaning to the buttons by doing things according to their role – starting and stopping the timer.

Finally, when this component is consumed by your page, only one listener is available – ‘onTimerCountdown()’. Users will interact with the timer, and when the timer counts down to 0, the listener you registered will be notified. You should be able to expect events at the right semantic level from all reusable components, from the simplest calendars to large complex components.

If a component can be made part of a larger document, two things you will care the most is serialization and dirty state. When users interact with the component and make a modification, you want to be told that the component is changed. This should trigger the dirty state of the parent. When the user clicks ‘Save’, you should be able to serialize the component and store this state in the larger document. Inversely, on bootstrap you should be able to pass the serialized state to the component to initialize itself.

Note that the actual technology used does not matter here – even the components embedded using iframes can use window.postMessage to send events up to the parent (and accept messages from the parent). While components living in your DOM will resize automatically, iframe-ed components will need to also send resizing events via window.postMessage to allow the parent to set the new size of the iframe.

The long tail

More complex reusable components don’t only have client-side presence. They have a need to call back to the server and fetch the data they need. You can configure such a component in two ways:

  1. You can fetch the data it requires for the component. In that case, the component is fully dependent on the container and it is container’s responsibility to perform all the XHR calls to fetch the data and pass it to the component. This approach may be best for pages that want full control of the network calls. As an added bonus, you can fit such a component into a data flow such as Flux, where some of the data may be coming from Web Socket driven server-side push, not just XHR request.
  2. You can proxy the requests that the component is performing. This approach is also acceptable because it allows the proxy to control which third-party servers are going to be whitelisted.
  3. You can configure CORS so that the component can make direct calls on its own. This needs to be done carefully to avoid the component siphoning data from the page to servers you don’t approve.

On all of these cases you may still want to be told about the events inside the web component using the component events as discussed above.

Frameworks are just the beginning

So there you go – all the problems you need to wrestle with when trying to reuse components in a larger project. Chances are the component is written in the ‘wrong’ framework, but trying to make the component load in your page is only the beginning. Fitting the component into the page visually, figuring out what is happening in it events-wise, and feeding it data from the server is the real battle. Unless you are trying to load a calendar widget, this is where you will spend most of your time.

© Dejan Glozic, 2016

Socket.io for Tardy React Status Indicators

The White Rabbit, Wikimedia Commons
The White Rabbit, Wikimedia Commons

In one of my favorite movies “The Blues Brothers”, the wife of the trucker pit stop owner proudly exclaims that they play ‘both kinds of music – Country and Western’. Readers of my blog know I too tend to serve two kinds of articles. In think-pieces, I tend to pontificate on a topic that buzzes between my ears that week. In ‘hey, look at this code we wrote’, I share something we learned by doing. I am sure there is an audience for both kinds but if you are in the latter camp, this is your day.

I’m late, I’m really really late

The problem we had to solve recently was as follows: a React store contains a list of items. These items are rendered in a hierarchy of React components, with a ‘dummy’ React component rendering a card for each item. The item has a number of visual elements on it that can be easily rendered from the item’s properties, but it also contains a status indicator.


This status indicator was killing us: when we fetch the array of items via the XHR API call, we can get them fairly quickly. However, each item represents an entity that can be working or not working (and therefore have a green or red status indicator). Computing this status takes a while, needs to be done separately for each item, and cannot hold the main XHR call.

We solved this problem with the help of the Socket.io module (at this point I need to remind you we use Node.js for our server side). You can apply the solution even if you are not using it, nor using Node at all – just use whatever Web Sockets library comes with your stack of choice.

The devil is in the pudding

OK, on to the details. What we do is as follows:

  • Whenever we want to update the status of the store, we issue an XHR.
  • The XHR fetches the items on the server via a downstream API call.
  • Before sending the response, the server endpoint initiates an array of async parallel calls to obtain status for each of the items. It sends the response to the client without waiting.
  • The client renders the cards, with the status rendered as a gray circle.
  •  When the status API calls return, the result is sent back via Web Socket channel as an event.
  • On the client, this event is converted into a Flux action, which in turn updates the Store, which updates the views.


Produce the code

OK, enough architecture, produce the code. In the interest of space conservation, I am going to assume you have rudimentary knowledge of both React and Flux. If not, Google away until React components, dispatcher, actions and stores are all concepts that make sense to you.

Our action creator is capable of dispatching two actions – fetching items and initiation service status update:

var AppDispatcher = require("./dispatcher");

var ServiceConstants = require("./action-constants");

var ServiceActions = {
	fetchItems: function(data) {
			actionType: ServiceConstants.FETCH_ITEMS,
			data: data
	serviceStatusUpdate: function(data) {
			actionType: ServiceConstants.SERVICE_STATUS_UPDATE,
			data: data

module.exports = ServiceActions;

OK, on to the ItemStore. It is a fairly standard affair. When asked to fetch items, it makes an XHR request to the endpoint in the Node.js server that returns server items. When status update arrives, it merges the provided status with the items and fires a change event again. This will drive the view to re-render:

// Modules

var Constants = require("./action-constants");
var EventEmitter = require("events").EventEmitter;

// Globals

var AppDispatcher = require("./dispatcher");
var XhrUtils = require("./xhr-util");

var EVENT_CHANGE = "items";
var items;
var emitter = new EventEmitter();
var listeningForSocket = false;

// Public Methods ------------------------------------------------------------->

var ItemStore = {

	getItems: function(type) {
		return items;

	emitChange: function(data) {
		emitter.emit(EVENT_CHANGE, data);

	addChangeListener: function(callback) {
		emitter.on(EVENT_CHANGE, callback);

	removeChangeListener: function(callback) {
		emitter.removeListener(EVENT_CHANGE, callback);


module.exports = ItemStore;

// Private Methods ------------------------------------------------------------>

function fetchItems(url) {
	if (!url) {
		console.warn("Cannot fetch items - no URL provided");

	// Fetch content
	XhrUtils.doXhr({url: url, json: true}, [200], function(err, result) {
		if (err) {
			console.warn("Error fetching assets from url: " + url);

		items = result;

function serviceStatusUpdate(data) {
	for (var i in data) {
		var s = data[i];

function _updateStatus(s) {
	for (var i in items) {
		var item = items[i];
		if (item.id===s.id) {
			item.status = s.status;

// Dispatcher ------------------------------------------------------------>

// Register dispatcher callback
AppDispatcher.register(function(payload) {
	var action = payload.action;

	// Define what to do for certain actions
	switch (action.actionType) {
		case Constants.FETCH_ITEMS:
			return true;

	return true;


Finally, the view. It fires the action that fetches items upon mounting, and proceeds to render the initial state with the loading indicator. Once the items are ready, it re-renders itself with the server items in place. Finally, when the status update arrives, it just re-renders and lets React diff the DOM for changed properties. Which is why we love React in the first place:

var React = require('react');
var Router = require('react-router');
var ItemStore = require('../flex/item-store');
var ActionCreator = require('../flex/action-creator');
var ActionConstants = require('../flex/action-constants');

module.exports = React.createClass({
  getInitialState: function() {
    return { loading: true, items: [] };

  componentDidMount: function() {

  componentWillUnmount: function() {

  _handleAssetsChanged: function(type) {
    if(type === ActionConstants.STATE_ERROR) {
        error: "Error while loading servers"
    } else {
        loading: false,
        error: null,
        items: ItemStore.getItems() || []

  render: function render() {
    var loading;
    var items;
    var error;

    if (this.state.loading) {
      loading = (
         <div className="items-loading">
            <div className="LoadingSpinner-dark"></div>
    else {
      items = (
          {this.state.items.map(function(item) {
            var statusClass = "status-indicator";
            if (item.status==="active")
              statusClass+= " status-active";
            else if (item.status==="error")
              statusClass+= " status-error";
              statusClass+= " status-loading";
            return (
               <li className="service-card" key={item.id}>
                  <div className="service-card-status">
                  <div className={statusClass}/></div>
                  <div className="service-card-name">{item.title}</div>
                  <div className="service-card-type">{item.dist}</div>
    if (this.state.error) {
      error = (
         <div className="error-box">{this.state.error}</div>
    return (
       <div id='list'>
             This list shows currently available servers and their status (
             <span className="text-loading">gray</span> - loading,&nbsp; 
             <span className="text-active">green</span> - active,&nbsp; 
             <span className="text-error">red</span> - error)

The status update portion involves Socket.io module. When the status is being computed, it is being emitted via the Web Socket. Of course, for this example we are faking the delay, but you can imagine actual status API request being involved:

var io = require('socket.io')(server);

function emitStatus(items) {
  var status = [];
  for (var i in items) {
    var item = items[i];
    var s = {};
    s.status = (Math.random()<.5)?"active":"error";
    s.id = item.id;
    setTimeout(function(status) {
      io.emit("status", [status]);
    }, (Math.floor((Math.random() * 900) + 100)), s);

This status even is being picked up on the client, producing the status update action that is dispatched, picked up by the store and eventually causing the view to update:

  var socket = io.connect(location.origin);
  socket.on('status', function (data) {

Conclusion and commentary

One of the many realization we have come to using React in production is that one can get very far with views alone. A top-level view can make XHR requests, then pass the data down to child components as props. Child components can capture low-level events, and translate them into events of higher level of abstraction via callbacks passed in as props.

We realized many simple pages can be entirely done this way. What we have noticed though is that when Web Sockets and server-side events are involved, Flux architecture really comes to its own and shows its potential. Mixing server side events with events produced by users interacting with the UI is much better done using Flux.

The problem we tried to solve is rendering the initial page. If the user stays on the page for a prolonged period of time, status changes of the items can continue, but the code above can handle it without any changes. Of course, if others can create or remove server items, new events and actions are needed to model that.

As usual, the entire example is available as a Node.js app on Github, and I have deployed it in Bluemix so that you can see the app running here: https://react-status-demo.mybluemix.net. Just keep refreshing the browser to see the three-step rendering (page, then items, and finally status).

© Dejan Glozic, 2016

Self-Contained Systems – Microservices for Dummies

Harran beehive houses, Wikimedia Commons
Harran beehive houses, Wikimedia Commons

I love the smell of the new year in the morning. As in the tweet above, there is something hopeful in starting a new blog by first updating the copyright notice at the bottom.

Nevertheless, what is currently occupying my mind is not as much how 2016 started, but how 2015 ended. As luck would have it, within weeks I had a chance to present about our production experience with micro services to two enterprises –  a mutual fund and a bank. Both were curious about them and simultanously unsure as to how they would apply in their unique situation.

In such discussions, the topic of Netflix inevitably comes up, and I am not sure this is a good thing. Sure, I enjoyed watching The Social Network like the most of us, but I am not sure it taught me much about how to become Mark Zuckerberg. Certain experiences, approaches and stories are so intensely singular that their utility as a ‘how to’ manual is quite poor.

In the case of Netflix, I have noticed two characteristics that are not very useful in most scenarios where micro services are considered:

  1. Very few outfits need to scale up to a third of all Internet traffic
  2. Being the micro service pioneers, Netflix engineers wrote a lot of custom code that is impractical in most situations

Maybe we don’t need 1000 microservices?

The consequence of the fact that you don’t need to prepare to handle a third of all Internet is that the overhead equation skews towards having fewer large micro services, rather than many small ones. In fact, it is surprising what transpires when discussions go a bit deeper into the ‘why’ direction.

As you probably know, I am on the record claiming that micro services are more about future-proofing, solving organizational problems of large teams and DevOps than any particular technology. It has become apparent to us as we talk to companies that the problem of two speed IT is more acute and pressing than replicating Netflix’s improbable feat. In the two-speed systems, companies can preserve investment and control through exposing the current system of record through APIs, while building up a cloud-based micro service system that consumes those APIs and serves Web and mobile experiences at the required speed and agility. You don’t need thousands of micro services for this.

In fact, more proof that the pragmatic pendulum has swung towards more reasonable architectures  can be found in the coining of a new phrase – Self Contained Systems (SCS). It is a neat workaround for the cognitive issues that may come from the term ‘micro service’. The term ‘system’ implies something larger, so even the illustrations of SCSs imply no more than a dozen or so. All the good properties of micro services are preserved (polyglot persistence, ability of teams to develop and deploy them independently, stack independence), but there is no expectation to deploy them in the hundreds and thousands. Sure, a self-contained system can in turn consist of several micro services, but at that point it becomes implementation detail.

It is probably best for companies dipping their toes into the micro service world to think about self-contained systems instead, lest they are taken down the Netflix path which will most likely be wrong for them.

I make my own underwear

The point of custom code from above reminded me of a story from the days when I was still climbing the audiophile ladder and cared about hi fi companies and stories about them. In this particular story from 2011, a parent company IAG that owns illustrious British names such as Quad, Mission and Wharfedale, describes its process like so:

One of the immediately notable things about IAG’s operation is the extent of what they make themselves: from new parts, to the tools to make those parts, through to all their own speaker drivers: the vast bulk of what goes into IAG products is made right there in the Shenzhen factory. They even manufacture their own wire.

What Hi Fi: Behind the scenes at IAG

The advice ‘start by making your own wire’ would make sense for very few companies, and the same applies to most enterprises trying to replicate Netflix’s approach. This comes into sharp focus when you start discussing concepts such as service discovery. Consider the following diagram:


The URL of the micro service system is composed based on spaces (dev, QA, prod) and regions. It can be easily constructed or scripted. The routing proxy partitions the path namespace, and each path acts as API and is driving a load balancer which in turn handles any number of micro service instances.

I am showing you this because we are running this successfully in production right now. If you want to, say, hit the development version of the micro service system running in US1 data centre, you would have something like https://foo-dev.us1.acme.com. A script with a couple of substituted variables is all you need to hit the system you want. As for the load balancer, it comes for free in the IBM’s Cloud Foundry-based PaaS (Bluemix).

If you have a Netflix-type system, you need service discovery just so that you can connect all your micro services, all flailing around at the unguessable IPs and port numbers. And since you need to build everything yourself, a simple thing such as load balancing that comes out of the box in all PaaSes and a number of IaaSes is something you need to provide.

Netflix story is unique for many reasons, and having to build everything yourself is one of those characteristics not necessary in 2016. Cloud today is all about what you can outsource to a PaaS or IaaS vs what is uniquely your value add. Building load balancers is doable but a rarely justified practice today.

For Dummies?

Yes, I know, forgive me for the click bait – SEO and all that. But by now you are well versed with the tongue-in-cheekness of the ‘For Dummies’ franchise. As people finding value in them are hardly dummies, so are people wisely choosing Self-Contained Systems instead of going down the path of the thousand micro services. Running a half a dozen to a dozen independently deployed, agile self-contained systems hooked up to your legacy systems may be the best choice for you right now. You can always add more, but don’t feel pressured to do so.

© Dejan Glozic, 2016

Dust.js as a React Delivery Vehicle


Even people in love with Single Page Apps and pooh-poohing server side rendering know that SPAs cannot just materialize in the browser. Something has to deliver them there. Until recently, AngularJS was the most popular client side JS framework, and I have seen all kinds of ADVs (Angular Delivery Vehicles) – from Node.js (N is for Node.js in the MEAN stack), to wonderfully ironic JSP files serving Angular. Particularly exuberant developers go all the way to using static HTML to launch their apps into orbit, extolling the virtues of CDNs.

Of course, I would like to see their faces the first time they realize the need to compute something dynamically in those static files. But that is beside the point. The important takeaway is that AngularJS has a very nice property of being embeddable. You don’t need to surrender your entire page to Angular – you can sequester it to a DIV (in fact, you can have multiple Angular apps running in the same page). This is a very nice feature, but also a necessary one – again, AngularJS is a pure client side framework.

With great power…

React changes this equation because it can be rendered on the server. Similarly to AngularJS, React root component can be mounted to any node inside your page, but unlike Angular, this page can also be rendered on the server by React.

This opens the door for a whole new class of isomorphic or universal apps, and I have written about it already. What you need to decide is whether you want to render the entire page, or a subset of it. If you choose the latter, then the subsequent question becomes – how do you start the rendering process until React takes over?

My first inclination was to go full React. Using the awesome react-engine module, it is easy to configure React as the view engine in the Express framework, and render the entire page. However, we soon hit some snags:

  1. As we render our UIs using micro services that use UI composition to glue the pages together, we hit a problem of React being grumpy about having to render HTML snippet arriving from outside of its control – something much easier with a templating library such as Dust.js.
  2. JSX turned out to be fairly quirky and finicky when rendering HTML HEAD element, as well as trying to inline JavaScript in tags. In fact, the latter became so error prone that we completely gave up on it, electing to put all our script in files. I didn’t like this particular restriction, and this remains an area where I am sour on JSX and the fact it most decidedly isn’t HTML.

Of course the problems I listed above did not show up in the small example I published in the initial article. They reared their ugly head once we started getting serious. In my original example, the header was a React component. This works great if you control the entire app, but the whole deal of UI composition is to integrate with apps not necessarily written using React, and you can only do that using the least common denominator (HTML, CSS and vanilla JS).

Remind me again about UI composition

In a nutshell, UI composition is the approach where common areas on the page are served as HTML snippets from a REST API. The API service can also serve CSS and JavaScript files that accompany the snippet (if you don’t serve them from a CDN). The idea is that a dedicated microservice can be providing these common areas this way, allowing other microservices to pull the content and inline it in their pages regardless of their stack.

UI composition inverts the flow you would normally use with iframes. Instead of rendering the common areas, then placing an iframe for the content and loading the content from an external URL, we load the common area as HTML and inline it into our page. In the olden days this approach was often accomplished using ‘edge side includes’ or ESIs. This approach has many benefits, including no need for the dreaded iframes, full control over the entire page, and the ability to integrate microservices implemented using different stacks. For example, the teams in my current project use:

  1. Node.js microservices using React for isomorphic rendering
  2. Node.js microservices using Dust.js and jQuery
  3. Java microservices using AngularJS

All of these microservices render the exact same header even though they use different stacks and libraries. Powerful stuff.

OK, so React?

The problem here with React is that when we use it to render the entire page server-side, it starts to squirm in discomfort. Inlining HTML is a problem for React because, while it can be used on the server, it was designed to shine on the client. Inlining raw HTML presumably obtained as a user input into DOM is a major security exposure without sanitization, and React will tell you so in no uncertain terms. Here is the line that inlines the shared header taken straight out of our production code:

<div dangerouslySetInnerHTML={{__html: decodeURIComponent(this.props.header)}} />

This is React telling us very clearly: “I really, really, really don’t like what you are asking me to do”. And this is even before we bring in react-engine. React engine is a PayPal module that enables isomorphism when combined with Node.js/express and react-router. We can render on the server, then pack the properties as part of the HTML response, hydrate the React components on the client and continue where we left off.

Only we have a little problem: notice that in the snippet above we have inlined the header by passing it down to the React component as a prop. What this means is that this prop (‘header’) will be sent to the client alongside other properties. The only problem is that this property contains HTML snippet for the entire header, and because it can really mess up react-engine, it is also URLencoded, making it even bulkier.

This is an unhappy situation: first we inline the header HTML (which is fine), then we send that same inlined HTML, this time encoded, as a React prop to the client. This unnecessarily bloats the generated HTML.


This is what we are going to do to solve this problem: we are going to call back our trusty Dust.js sitting sad on the sidelines ever since we got smitten with React. We can press it into service to do what it does best – outline the page, prep it up to the point where React can take over.

It’s a good thing that Express framework will happily handle more than one view engine. Of course, only one of them can be registered as default, but this just means that we will need to spell out the extension of dust files (because React is the default). Big deal.

var path = require('path')
, express = require('express')
, renderer = require('react-engine')
, dust = require('dustjs-linkedin')
, helpers = require('dustjs-helpers')
, cons = require('consolidate');


// create the view engine with `react-engine`
var engine = renderer.server.create({
  routes: require(path.normalize(__dirname + '/public/routes.jsx')), 
  docType: ""

// set the engines
app.engine('.jsx', engine);
app.engine('.dust', cons.dust);

// set the view directory
app.set('views', __dirname + '/public/views');

// set jsx as the view engine
app.set('view engine', 'jsx');

// finally, set the custom view
app.set('view', renderer.expressView);

Once we have both engines set up, we can arrange our delivery vehicle like this. We will first set up a reusable layout template using Dust.js that will deliver the outline of each page:

<!DOCTYPE html>
   <meta charset='utf-8'>
   <meta http-equiv="X-UA-Compatible" content="IE=edge">
   <meta name="viewport" content="width=device-width, initial-scale=1" >
   <link rel="stylesheet" href="/css/styles.css">
   <div class="main-content">

Notice how inlining of the header now happens in Dust.js, relieving React from doing it. We can now use this template to render our page like this:

<div id="react-mount-node">{spa|s}</div>
<script src="/bundle.js"></script>

We have created a DIV in the template above where we are mounting React, and we will inline the React server-side content using the ‘spa’ variable.

Our Express controller that will handle both Dust.js and React can look something like this:

var request = require("request");

module.exports.get = function(req, res) {
  var settings = {
    title: 'SPA - Demo',
    name: 'React SPA',
    selection: 'header-spa'

  var headerUrl = "http://"+req.headers.host+"/header?selection=header-spa";

  request.get(headerUrl, function (err, response, body) {
    if (err)
      settings.header = body;
    res.render(req.url, { name: settings.name }, function(err, html) {
      if (err) {
        settings.spa = err.message;
        settings.spa = html;
      res.render("spa.dust", settings); 

The code above is what I call ‘React Delivery Vehicle’. It has three steps:

  1. First we fetch the header from the URL that is providing us with the header HTML snippet. We capture it into a variable ‘header’. In production, we will heavily cache this step.
  2. Then we render the React root component as usual. This will employ react-engine and react-router to render all the views necessary for the provided request URL. However, we don’t send the rendering to the response. Instead, we capture it in the variable ‘spa’.
  3. Finally, we invoke Dust.js view engine to render the full page, passing in the ‘settings’ object. This object will contain both the header and the content rendered by React. Dust will simply inline both while rendering the page outline. The result will be sent directly to the Express response.


The solution described above plays to the strengths of both Dust.js as a server side template engine and React as a component library. Dust.js is great at rendering HTML outline the way you would expect, with no need to worry about JSX quirks around HEAD meta tags, or JavaScript inlining. React takes over to render isomorphic components into the mount node. This means we don’t need to send any data that is not needed on the client. This fixes the HTML bloat problem I mentioned before.

As for the negatives, setting up two rendering engines on the server has a slight overhead, and switching mentally from Dust.js to JSX adds context switching tax. Luckily, you can set up reusable Dust.js templates and not really worry about them too much – most of the action will be in JSX anyway.

We like this approach and are currently switching to it across the board. If you have alternative ideas or comments, drop me a line. Meanwhile, the source code for the entire example is available on GitHub as usual, and the sample app employing this mechanism runs on Bluemix.

© Dejan Glozic, 2015