Spring Boot and Apache ZooKeeper: Dynamic Config, Service Discovery, and Leader Election

I have been building distributed systems for over a decade, and I still remember the mess we made when we tried to manage configuration across forty microservices by hand. We relied on environment variables, property files packed into Docker images, and frantic Slack messages every time a database URL changed. That approach broke regularly. Then I discovered Apache ZooKeeper, and later, how to plug it into Spring Boot. That combination saved my sanity. Let me walk you through why it matters and how you can do it, step by step.

Think about the last time you updated a configuration value. Did you have to restart the entire application? If you are using Spring Boot alone, the answer is yes. That is because properties are loaded once at startup. In a distributed setup, restarting ten instances sequentially can take minutes, cause downtime, and lead to inconsistency between nodes. ZooKeeper solves this by acting as a central coordination service. It stores configuration data in a tree structure called a “znode”. Your Spring Boot application can watch that znode and pick up changes the moment they are written.

So how do you integrate them? The official Spring Cloud ZooKeeper project provides a simple way. You add the dependency to your pom.xml:

<dependency>
    <groupId>org.springframework.cloud</groupId>
    <artifactId>spring-cloud-starter-zookeeper-config</artifactId>
</dependency>

Then you tell Spring Boot where ZooKeeper lives. In bootstrap.properties:

spring.cloud.zookeeper.connect-string=localhost:2181
spring.cloud.zookeeper.config.enabled=true

Now any configuration stored under the path /config/myapp/datasource can be read as a property. For example, if you create a znode at /config/myapp/datasource/url with value jdbc:mysql://newhost:3306/db, you can use @Value("${datasource.url}") in your code. The magic is that when that znode changes, Spring Cloud can refresh the context automatically using @RefreshScope. This means no restart, no redeploy, no panic.

But wait — does your application know how to react to a change? That is the tricky part. You must annotate any bean that holds configuration with @RefreshScope. Otherwise, the old value stays cached. I once spent two hours debugging why my database connection pool wasn’t updating. The bean wasn’t refreshable. Always add @RefreshScope to any component that depends on dynamic properties.

@Component
@RefreshScope
public class DatabaseConfig {
    @Value("${datasource.url}")
    private String url;

    public void printUrl() {
        System.out.println("Current URL: " + url);
    }
}

Now, what about coordination between services? Have you ever needed to ensure only one instance of a scheduled task runs across a cluster? That is where leader election comes in. ZooKeeper makes this dead simple. You can use the Curator framework (an Apache project that wraps ZooKeeper) inside Spring Boot.

First, add Curator:

<dependency>
    <groupId>org.apache.curator</groupId>
    <artifactId>curator-recipes</artifactId>
    <version>5.5.0</version>
</dependency>

Then create a LeaderLatch bean:

@Bean
public CuratorFramework curatorFramework() {
    return CuratorFrameworkFactory.newClient("localhost:2181", new RetryOneTime(1000));
}

@Bean
public LeaderLatch leaderLatch(CuratorFramework client) throws Exception {
    LeaderLatch latch = new LeaderLatch(client, "/leader-election");
    latch.start();
    return latch;
}

Now in your scheduled task, check if this instance is the leader:

@Scheduled(fixedRate = 10000)
public void runIfLeader() {
    if (leaderLatch.hasLeadership()) {
        System.out.println("I am the leader! Doing the job.");
        // perform critical task
    } else {
        System.out.println("Not the leader, skipping.");
    }
}

This pattern prevents duplicate database cleanup jobs or cache refreshes. I used it on a project where we had five instances of a report generator. Without leader election, each one would generate the same PDF, wasting CPU and confusing users. With this code, only one instance runs the report.

Here is a personal touch: the first time I integrated ZooKeeper with Spring Boot, I forgot to add the bootstrap.properties file. I spent three hours looking at logs. The error message was cryptic: “Could not resolve placeholder”. I had defined the property in ZooKeeper, but Spring Boot wasn’t even looking there. The bootstrap file is essential because it triggers the initial ZooKeeper connection before the main application context loads. Without it, your dynamic properties won’t exist.

Now I want you to pause and ask yourself: how often do you restart your application just to change a log level? If you are like most developers, it happens more than you want. With ZooKeeper, you can change the logging level for all instances instantly. You can even toggle feature flags without a deploy.

But there is a subtlety: ZooKeeper is not a configuration database in the traditional sense. It is a coordination service. You should not put large blobs of data in znodes — they have file size limits. For big configurations, store them in a dedicated database and use ZooKeeper just to hold the pointer. Also, ZooKeeper transactions are not optimized for high write throughput. If thousands of config changes happen per second, consider another tool. But for most enterprise use cases, it is perfect.

Another use case I want to mention is service discovery. Spring Cloud ZooKeeper also provides service registration and discovery out of the box. When a Spring Boot application starts, it can register its address in ZooKeeper under a service name. Other applications can then look up the service by name without hardcoding IPs. This is a lifesaver when you scale instances up and down dynamically.

To enable discovery, just add:

<dependency>
    <groupId>org.springframework.cloud</groupId>
    <artifactId>spring-cloud-starter-zookeeper-discovery</artifactId>
</dependency>

Then use @EnableDiscoveryClient on your main class. Spring Cloud will automatically register the service. Another service can call it using RestTemplate with @LoadBalanced. No manual address management.

So why should you care? Because distributed systems are not optional anymore. Even small projects now run multiple instances across containers. Coordination becomes a pain without a central brain. ZooKeeper, married with Spring Boot, gives you that brain without reinventing the wheel.

I learned this the hard way. In my early career, I tried to build custom leader election with a database table. It had race conditions, and we ended up with three leaders at the same time, corrupting data. The fix took weeks. ZooKeeper’s consensus protocol (Zab) guarantees only one leader, so that problem disappears.

Now let me ask you: how many times have you wished you could change a configuration in production without touching any running process? If you integrate ZooKeeper today, you can do that tomorrow.

Before you leave, I encourage you to try this integration in a small project. Set up ZooKeeper locally (it runs as a single node for testing), create a sample Spring Boot app, and make a dynamic property change. Watch the log show the updated value without a restart. It feels like magic, but it is just good engineering.

If this article helped you understand how Spring Boot and ZooKeeper work together, please like it. Share it with a colleague who still restarts services for every config change. And leave a comment below about your own experiences with distributed coordination. I read every comment and would love to hear how you solve these problems in your environment.

Now go make your systems a little more resilient. You have the tools.

---

As a best-selling author, I invite you to explore my books on Amazon. Don’t forget to follow me on Medium and show your support. Thank you! Your support means the world!

---

101 Books

101 Books is an AI-driven publishing company co-founded by author Aarav Joshi. By leveraging advanced AI technology, we keep our publishing costs incredibly low—some books are priced as low as $4—making quality knowledge accessible to everyone.

Check out our book Golang Clean Code available on Amazon.

Stay tuned for updates and exciting news. When shopping for books, search for Aarav Joshi to find more of our titles. Use the provided link to enjoy special discounts!

---

📘 Checkout my latest ebook for free on my channel!
Be sure to like, share, comment, and subscribe to the channel!

---

Our Creations

Be sure to check out our creations:

Investor Central | Investor Central Spanish | Investor Central German | Smart Living | Epochs & Echoes | Puzzling Mysteries | Hindutva | Elite Dev | JS Schools

---

We are on Medium

Tech Koala Insights | Epochs & Echoes World | Investor Central Medium | Puzzling Mysteries Medium | Science & Epochs Medium | Modern Hindutva