Apache Spark / Big Data Administrator

Job Description

• 10+ years solid experience with deploying, configuring, monitoring, and troubleshooting Apache Spark clusters and associated big data frameworks (such as Hadoop, Kafka, or cloud-based data platforms) to ensure high availability and performance.
• Focuses on cluster health, performance tuning, and production data pipelines.
• Technical Experience: Deep hands-on experience with Apache Spark architecture (Driver, Workers, Executors, RDDs/DataFrames) and cluster managers (YARN, Kubernetes, or Mesos).
• OS & Scripting: Strong proficiency in Linux/Unix administration and scripting languages (Bash, Python, or Scala).
• Ecosystem Knowledge: Familiarity with big data and streaming tools like Hadoop HDFS, Kafka, Hive, or monitoring suites (Splunk, Prometheus, Grafana).
• Debugging Skills: Strong analytical capabilities to inspect logs, thread dumps, and metrics under heavy production loads
• Excellent coordination with stakeholders and communication skills

Must have experience with below responsibilities

• Cluster Management: Install, configure, and maintain Apache Spark clusters and node infrastructure in production and staging environments.
• Performance & Monitoring: Monitor cluster health, resource utilization (CPU, memory), and execution logs using tools like Spark UI, Splunk, or Elastic. Identify bottlenecks and optimize job execution times.
• Troubleshooting: Diagnose and resolve out-of-memory errors, executor failures, shuffle issues, and node crashes in distributed workloads.
• Pipeline & Integration Support: Support data engineers and developers in managing Spark applications, streaming jobs, and integrations with message brokers like Kafka.
• Security & Access Control: Implement security principles, role-based access control, authentication, and data encryption across the cluster.
• Backup & Disaster Recovery: Execute backup, recovery, and failover plans for Spark applications, configurations, and critical data stores.

Patching & Upgrades: Apply software patches, driver updates, and version upgrades with minimal disruption to ongoing production jobs.