Blog
Thoughts on DevOps, database engineering, and the art of automation.
Big Data
Scaling HDFS: Fixing the "Protocol Message Too Large" Block Report Error
When DataNodes send Full Block Reports exceeding 64 MB, the NameNode rejects them. Here's how to fix the ipc.maximum.data.length limit in large Hadoop clusters.
Mar 30, 20262 min read
Big Data
Understanding the Relationship Between HDFS fsimage and Cloudera Reports Manager Heap Size
If Cloudera Manager's Reports Manager is consuming high CPU or memory, the root cause is often improper heap sizing relative to your HDFS namespace size (fsimage).
Mar 14, 20263 min read
Big Data
How to Solve the Hive Small File Problem in Hadoop (Practical Guide for Production Clusters)
10,000 files × 1 MB is far worse than 100 files × 128 MB for the same total data. Here's how to detect, prevent, and fix the small file problem in production Hive clusters.
Mar 9, 20265 min read
Big Data
How to Configure a Read-Only Trino User for Hive Tables (File-Based Access Control)
Grant analysts and BI tools read-only access to Hive tables in Trino without Apache Ranger — using file-based access control and HDFS ACLs.
Mar 7, 20263 min read