The WiredTiger write-ahead log is a critical component of MongoDB's storage engine, responsible for ensuring data durability in the event of a system crash. The design was optimized for high-performance and low-latency writes, but when MongoDB became the default storage engine, it faced challenges due to changes in usage patterns. A bottleneck, known as the "logjam," emerged due to two condition mismatches: one related to the durability level used by clients (no-sync vs. write-only) and another related to the number of threads versus cores available. These mismatches led to priority inversion, causing threads to busy-wait on state changes, consuming CPU cycles. A new approach was needed, which involved rethinking assumptions and optimizing for the current conditions. The solution involves a stand-alone prototype that uses atomic operations to manage slot state and buffer writes, achieving significant performance improvements in the "no-sync" case.