0 views collected around this technical thread.
This article walks through a real‑world case where a simple INSERT in TiDB runs for tens of seconds, explains how transaction modes and write conflicts trigger SQL retries, and outlines how to diagnose and configure retry behavior to avoid slow writes.
This article compares TiDB's optimistic and pessimistic transaction models through a real‑world high‑concurrency write‑conflict scenario, showing how pessimistic locking can achieve over ten times higher throughput than optimistic retries.
This article analyzes a TiDB cluster performance incident where QPS dropped and duration spiked due to write‑write conflicts, detailing the monitoring data, root‑cause investigation of server‑busy and scheduler latch issues, and the attempted mitigation steps such as enabling txn‑local‑latches and adjusting insert statements.
This article analyzes a TiDB production cluster outage caused by write‑write conflicts, walks through the monitoring data and log investigation, explains the underlying transaction model differences, and shares the step‑by‑step troubleshooting process that led to the final resolution.