In this section, we compare the performance of eager writing against that of update-in-place and logging with a variety of micro-benchmarks. We first run the small file and large file benchmarks that are commonly used by similar file system studies. Then we use a benchmark that best demonstrates the strength of eager writing: small random synchronous updates with no idle time, which also illustrates the effect of disk utilization. Next we examine the effect of technology trends. Last, we examine how the availability of idle time impacts the performance of eager writing and logging. Unless explicitly stated to be otherwise, the experimental platform is the SPARCstation-10 running on the simulated Seagate disk.