https://akmalariq.dev/blog/ Notes on data engineering, storage, and system design — written to be re-read. en https://akmalariq.dev/blog/late-arriving-data/ https://akmalariq.dev/blog/late-arriving-data/ Tue, 13 Oct 2026 00:00:00 GMT Event time is not arrival time. How late data breaks closed reports and incremental loads, and the patterns that make it harmless: watermarking, lookback reprocessing, idempotent partitions, and versioned event logs. streaming event-time watermarks idempotency data-quality https://akmalariq.dev/blog/break-your-cdc-pipeline/ https://akmalariq.dev/blog/break-your-cdc-pipeline/ Tue, 06 Oct 2026 00:00:00 GMT A lab that kills the consumer mid-batch, poisons the topic, and replays a consumer group — plus the bug this pipeline shipped with and how I found it. kafka dlq observability testing streaming https://akmalariq.dev/blog/effectively-once-cdc-clickhouse/ https://akmalariq.dev/blog/effectively-once-cdc-clickhouse/ Tue, 29 Sep 2026 00:00:00 GMT Kafka only promises at-least-once. Here is how write-then-commit, source-timestamp versions, and ReplacingMergeTree make duplicates harmless — without claiming exactly-once. clickhouse kafka idempotency exactly-once streaming https://akmalariq.dev/blog/change-data-capture-explained/ https://akmalariq.dev/blog/change-data-capture-explained/ Tue, 22 Sep 2026 00:00:00 GMT Why a nightly sync is never enough — and how Debezium, Kafka, and a Python consumer turn the write-ahead log into a live replica of your database. cdc streaming debezium kafka postgres https://akmalariq.dev/blog/file-vs-block-vs-object-storage/ https://akmalariq.dev/blog/file-vs-block-vs-object-storage/ Wed, 16 Sep 2026 00:00:00 GMT The differences that actually matter to a data engineer — and why your lakehouse lives on objects while your database lives on blocks. storage fundamentals system-design