Worked on the grafana/tempo repository to enhance observability and debugging by adding block size information to log messages when head blocks are cut. This backend development effort, implemented in Go, improved visibility into memory pressure and flush operations by providing clearer context in logs during high-load scenarios. The approach involved standardizing log messages across both the ingester and livestore components, supporting unified tracing and troubleshooting. By instrumenting these areas, the work laid the groundwork for future proactive monitoring, such as alerts and dashboards focused on block size and memory events, ultimately reducing time-to-insight for operational issues.