<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>navneetlal — writing</title><description>Field notes on distributed systems, data platforms, and large-scale migrations.</description><link>https://navneetlal.in/</link><language>en-us</language><item><title>Migrating 45 services from EC2/systemd to EKS — and cutting deploys 4.5×</title><link>https://navneetlal.in/writing/migrating-45-services-to-eks/</link><guid isPermaLink="true">https://navneetlal.in/writing/migrating-45-services-to-eks/</guid><description>A field report from leading a platform migration of 45 services and 250+ instances at a multi-tenant SaaS company.</description><pubDate>Thu, 14 May 2026 00:00:00 GMT</pubDate><category>Migrations</category><category>EKS</category><category>Kubernetes</category><category>SRE</category></item><item><title>Why we replaced analytics on Mongo with StarRocks</title><link>https://navneetlal.in/writing/replacing-mongo-analytics-with-starrocks/</link><guid isPermaLink="true">https://navneetlal.in/writing/replacing-mongo-analytics-with-starrocks/</guid><description>What broke when analytical queries shared a tenant cluster with OLTP — and what moving to a real OLAP engine actually gave us.</description><pubDate>Tue, 28 Apr 2026 00:00:00 GMT</pubDate><category>Data Platforms</category><category>StarRocks</category><category>Kafka</category><category>Analytics</category></item><item><title>Cascading dep hell: Java 8 → 17 across 35 services</title><link>https://navneetlal.in/writing/java-8-to-17-cascading-dep-hell/</link><guid isPermaLink="true">https://navneetlal.in/writing/java-8-to-17-cascading-dep-hell/</guid><description>End-of-support runtimes don&apos;t just block CVE fixes — they block every transitive update that depends on them. A field report from a 35-service upgrade.</description><pubDate>Fri, 10 Apr 2026 00:00:00 GMT</pubDate><category>Migrations</category><category>Java</category><category>Dependencies</category></item><item><title>Building a CDC pipeline with Debezium — without breaking the source</title><link>https://navneetlal.in/writing/cdc-pipeline-with-debezium/</link><guid isPermaLink="true">https://navneetlal.in/writing/cdc-pipeline-with-debezium/</guid><description>Lessons from tailing a multi-tenant Mongo oplog into Kafka without putting the source database under more load than it already had.</description><pubDate>Fri, 20 Mar 2026 00:00:00 GMT</pubDate><category>Distributed Systems</category><category>Debezium</category><category>Kafka</category><category>Mongo</category><category>CDC</category></item><item><title>Normalizing GPS telemetry from 8k vehicles across half a dozen sources</title><link>https://navneetlal.in/writing/normalizing-gps-from-8k-vehicles/</link><guid isPermaLink="true">https://navneetlal.in/writing/normalizing-gps-from-8k-vehicles/</guid><description>Polished APIs, flaky devices, customer-internal feeds — what cross-source GPS normalization actually looks like in production.</description><pubDate>Thu, 05 Mar 2026 00:00:00 GMT</pubDate><category>Distributed Systems</category><category>GPS</category><category>Telemetry</category><category>Logistics</category><category>Normalization</category></item><item><title>A lakehouse with Iceberg and Airflow — bronze, silver, gold</title><link>https://navneetlal.in/writing/lakehouse-with-iceberg-and-airflow/</link><guid isPermaLink="true">https://navneetlal.in/writing/lakehouse-with-iceberg-and-airflow/</guid><description>Why the three-tier lakehouse pattern works for replacing OLAP-on-OLTP, and what each tier actually does in practice.</description><pubDate>Wed, 18 Feb 2026 00:00:00 GMT</pubDate><category>Data Platforms</category><category>Iceberg</category><category>Spark</category><category>Airflow</category><category>Lakehouse</category></item><item><title>OpenTelemetry: trace, log, metric — one stack, three signals</title><link>https://navneetlal.in/writing/opentelemetry-stack-trace-log-metric/</link><guid isPermaLink="true">https://navneetlal.in/writing/opentelemetry-stack-trace-log-metric/</guid><description>What it actually costs to consolidate observability onto OTel + Grafana + Loki + Tempo + VictoriaMetrics, and what you get back.</description><pubDate>Mon, 02 Feb 2026 00:00:00 GMT</pubDate><category>Distributed Systems</category><category>OpenTelemetry</category><category>Observability</category><category>Grafana</category></item></channel></rss>