|
Description
|
About the Role
We are migrating a portfolio of Apache Kafka clusters from on-premises platforms (Cloudera and Confluent) to AWS MSK, managed through a CI/CD pipeline built on Terraform and GitLab. We are looking for a senior, hands-on Kafka engineer to own day-to-day Kafka administration across both environments, drive the migration, and coordinate closely with producer and consumer application teams to ensure a smooth, low-risk cutover.
This is a senior individual-contributor role for someone who has run Kafka at scale in production, is comfortable in a hybrid on-prem/cloud world, and can act as the technical point of contact between platform and application teams.
Required Qualifications
-
5 years in data/platform/infrastructure engineering, with 4 years operating Apache Kafka in production.
-
Hands-on administration experience with Cloudera and Confluent Platform.
-
Production experience with AWS MSK (provisioning, configuration, IAM auth, networking).
-
Proven experience migrating Kafka workloads to the cloud with minimal disruption.
-
Strong Terraform and GitLab CI/CD experience for infrastructure automation.
-
Hands-on Datadog experience for monitoring, dashboards, and alerting.
-
Deep understanding of Kafka internals: partitions, replication, ISR, offsets, consumer groups, and delivery semantics.
-
Solid AWS fundamentals: VPC, security groups, IAM, KMS, CloudWatch, PrivateLink.
-
Strong Linux, networking, and scripting skills (Bash, Python).
-
Excellent communication and coordination skills to work across producer/consumer teams.
Preferred / Nice to Have
-
Experience with MirrorMaker 2, Confluent Replicator, or cluster linking for replication/migration.
-
Schema Registry, Kafka Connect, and ksqlDB experience.
-
Kafka Streams or Flink familiarity.
-
Kubernetes / EKS experience (e.g., Strimzi).
-
Certifications: Confluent Certified Administrator, AWS Solutions Architect/DevOps.
-
Experience in a regulated or large enterprise environment.
|
|
Top Skills
|
Kafka Administration (On-Prem Cloud)
-
Administer and operate Kafka clusters across Cloudera, Confluent Platform, and AWS MSK.
-
Manage topics, partitions, replication, retention, quotas, ACLs, and consumer groups.
-
Tune brokers, producers, and consumers for throughput, latency, and reliability.
-
Handle patching, upgrades, capacity planning, and cluster health/performance troubleshooting.
-
Configure and maintain security: TLS/SSL, SASL, mTLS, IAM (for MSK), RBAC/ACLs, and encryption at rest and in transit.
-
Migration to AWS MSK
-
Plan and execute migration of clusters and workloads from on-prem (Cloudera/Confluent) to AWS MSK.
-
Design cutover strategies (e.g., MirrorMaker 2 / Confluent Replicator / cluster linking) with minimal downtime and no data loss.
-
Validate topic parity, offsets, throughput, and data integrity pre- and post-migration.
-
Define rollback plans and run migration dry-runs in lower environments.
CI/CD & Infrastructure as Code
-
Build and maintain Kafka/MSK infrastructure using Terraform.
-
Manage automated deployments and configuration through GitLab CI/CD pipelines.
-
Implement GitOps practices for topic, ACL, and cluster configuration management.
-
Enforce version control, code review, and environment promotion (dev ? test ? prod).
Observability & Monitoring
-
Build and maintain monitoring, alerting, and dashboards in Datadog for Kafka/MSK (broker metrics, consumer lag, partition health, JVM, throughput).
-
Establish SLAs/SLOs and proactive alerting to reduce incidents and mean-time-to-resolution.
-
Correlate Kafka metrics with application and infrastructure telemetry.
Producer / Consumer Coordination
-
Serve as the primary technical liaison between the Kafka platform team and producer/consumer application teams.
-
Advise application teams on client configuration, serialization/schema management, partitioning strategy, idempotence, and exactly-once/at-least-once semantics.
-
Support onboarding of new producers/consumers and troubleshoot client-side issues (rebalancing, lag, poison messages).
-
Coordinate migration schedules, testing, and cutover windows with each impacted team.
|
