Apache ZooKeeper with Spring Boot for Dynamic Config and Service Discovery

I remember the exact moment I realized I had been doing distributed configuration all wrong. I was staring at a production incident where a single database password change required restarting fifty microservices, one by one, at 2 AM. The deploy pipelines were queued, the operations team was sweating, and somewhere a cron job was failing because it couldn’t see the new endpoint. That night I swore I would never manually manage configuration files across a cluster again. That’s when I started looking deeply at Apache ZooKeeper and how it could be paired with Spring Boot to make these painful scenarios disappear.

Apache ZooKeeper is not a database. It’s a coordination service, a central brain for distributed systems that stores small pieces of critical state in a hierarchical tree. Think of it as a filesystem where each node can hold data and have children, but with built-in watches and consensus. When you combine it with Spring Boot, you get a way to push configuration changes instantly, discover services without hardcoded URLs, and coordinate tasks like leader election across instances. The magic is in the watchers. A service can subscribe to a ZooKeeper node, and when that node’s data changes, the service receives a notification and can reload its configuration without a single restart.

Let me show you how this works in practice. First, you add the Spring Cloud ZooKeeper dependency to your pom.xml:

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

Then, in your bootstrap.yml, you tell Spring Boot where to find ZooKeeper:

spring:
  cloud:
    zookeeper:
      connect-string: localhost:2181
      config:
        root: /config
        enabled: true

Now any property stored under /config/myapp in ZooKeeper becomes available as a Spring property. Change the value in ZooKeeper, and your application picks it up automatically if you use @RefreshScope. Have you ever had to email a configuration file around a team just to change a log level? With ZooKeeper, you change one node, and every instance sees it immediately.

But configuration is only half the story. Service discovery is where ZooKeeper really shines. Instead of hardcoding service URLs or using brittle environment variables, you register each Spring Boot instance as an ephemeral node. When the instance dies, the node disappears automatically. Other services can then query ZooKeeper to find healthy instances. Here’s how you enable it:

spring:
  cloud:
    zookeeper:
      discovery:
        enabled: true
        register: true

Then in your code, you can use DiscoveryClient to locate a service:

@Autowired
private DiscoveryClient discoveryClient;

public List<ServiceInstance> getInstances(String serviceId) {
    return discoveryClient.getInstances(serviceId);
}

This feels almost magical the first time you see it work. I remember testing this with three instances of a payment service running locally. I killed one, and within seconds the other two were the only ones returned by the discovery client. No load balancer configuration, no manual DNS updates. Just clean, dynamic discovery.

Now let’s talk about coordination. In distributed systems, there are tasks that only one instance should perform at a time, like purging old records or recalculating cache. ZooKeeper provides a primitive called leader election. With Spring Cloud and Apache Curator, this becomes straightforward. First, add the Curator dependency:

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

Then set up a leader latch:

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

@Scheduled(fixedRate = 60000)
public void runOnlyIfLeader() throws Exception {
    if (leaderLatch.hasLeadership()) {
        // perform the task
        System.out.println("I am the leader, running cleanup.");
    }
}

I once used this pattern to ensure that only one instance of a notification sender processed messages at a time in a clustered environment. Without it, we would have sent duplicate notifications to users. The leader election gave us correctness without complexity. What happens when the leader crashes? The latch automatically selects another instance. No human intervention needed.

Another essential coordination primitive is distributed locking. Imagine two microservices that both try to update the same file in a shared filesystem. Without a lock, you get corruption. ZooKeeper can act as a distributed mutex:

InterProcessMutex lock = new InterProcessMutex(client, "/locks/resource-update");
try {
    if (lock.acquire(10, TimeUnit.SECONDS)) {
        // critical section
    }
} finally {
    lock.release();
}

I’ve used this in scenarios where multiple services needed to atomically update a configuration value that was not suitable for ZooKeeper’s own nodes. The lock ensures that only one process writes at a time, preventing race conditions.

Now, you might ask: Is ZooKeeper still relevant in the era of Kubernetes and etcd? Absolutely. ZooKeeper has been battle-tested at massive scale by companies like Yahoo and Apache Hadoop. It offers strong consistency guarantees (sequential consistency) that many modern coordination stores do not. And with Spring Boot, the integration is so seamless that adding ZooKeeper feels like adding any other Spring dependency.

But there are gotchas. ZooKeeper is not a general-purpose database. It operates best when data fits in memory and changes infrequently. Don’t store large configuration blobs or frequent log data in it. Keep your nodes small. Also, connection strings and session timeouts require careful tuning depending on network latency. I recommend starting with a three-node ensemble for production, and monitoring the ZooKeeper’s zab state regularly.

When you start building with ZooKeeper and Spring Boot, you will notice a shift in how you think about distributed state. Instead of treating configuration and coordination as static, you start designing for change. Your services become self-configuring, self-healing, and much more resilient to failures.

I have personally seen teams reduce deployment times from hours to minutes after adopting this integration. The operational overhead of Syncing configuration files or managing service endpoints disappears. The system becomes a living organism that adapts to its environment.

So, what stops you from trying this today? You can set up a local ZooKeeper container with Docker in under a minute:

docker run --name zookeeper -p 2181:2181 zookeeper:3.8

Then bootstrap a simple Spring Boot project with the dependencies I showed. Change a property in ZooKeeper’s /config tree and watch your application respond. That moment of seeing log levels change in real time, without a restart, is when you’ll understand why this matters.

If you found this article useful, please like, share, and comment with your own experiences or questions about distributed coordination. I read every comment and I’d love to hear how you handle configuration and service discovery in your systems. Your feedback helps me write better, more practical content for developers like you.

---

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