Question about Kafka Managed Service – Connection Limits / Rate Limiting

Hi!

We are using the Managed Kafka service for a project with:

• >100 topics
• ~25 microservices producing and consuming messages concurrently

During a stress test, we observed that:

• With low concurrent produce/consume rates, everything works fine.
• When producers start producing at >50 events/sec per service, some services start disconnecting. Consumers and producers fail with errors like:

{"level":"ERROR","timestamp":"[REDACTED]","logger":"kafkajs","message":"[BrokerPool] Failed to connect to seed broker, trying another broker from the list: Connection error: read ECONNRESET","retryCount":0,"retryTime":280} {"level":"ERROR","timestamp":"[REDACTED]","logger":"kafkajs","message":"[Connection] Connection error: read ECONNRESET","broker":"kafka-***.ondigitalocean.com:****","clientId":"kafka-test-***","stack":"Error: read ECONNRESET\n at TLSWrap.onStreamRead (node:internal/stream_base_commons:218:20)"}

We don’t believe Kafka itself imposes these limits. Based on past experience with Redis, we suspect it might be related to:

• Firewall or network connection limits
• Managed service account limits

Questions:

1. Is there a rate limit on concurrent connections or message throughput for DigitalOcean Managed Kafka?

2. Are there recommendations to produce high volumes of events without causing service disconnections?

3. Could the issue be caused by network/firewall limits rather than Kafka itself?

4. If this is not a service limit, and the error is in our application (i.e., there is absolutely no limit on Kafka connections with private or public connection strings), how can we detect or fix the problem?

5. If this is a DigitalOcean service limit, how can we change or increase the concurrent connection limit?

We want to ensure our microservices can produce and consume high volumes of events without bottlenecking other services.

Thanks in advance for your guidance.