Pulumi recently introduced a significant performance improvement that can speed up operations by up to 20 times. Traditionally, Pulumi saves a snapshot of cloud infrastructure at each step of an operation, which ensures a current view of state even during failures. However, this approach slows down performance, especially for large stacks. The new feature, called journaling, enables faster operations by reducing the amount of data sent during these snapshots. Benchmark tests on two Pulumi projects illustrate the improvements. In the first test, creating a stack with over 3,000 AWS bucket objects took 58 minutes without journaling compared to just under 3 minutes with it. A similar pattern emerged when only half the resources changed, with times dropping from nearly 35 minutes to 1 minute and 45 seconds. Another test involving the setup of a Pulumi app and API showed a reduction from about 18 minutes to just over 9 minutes. These benchmarks were conducted using Pulumi Cloud in AWS’s us-west-2 region. The article provides an overview of how Pulumi's snapshotting works. Each operation starts and ends with a snapshot to track resource changes, which involves creating a pending operation in the snapshot until the operation is complete. However, the process of serially uploading these snapshots has been a bottleneck, particularly for large stacks. To address this, Pulumi implemented a diff-based protocol that only sends the changes between the old and new snapshots, reducing data transfer and improving performance. Despite these advancements, snapshotting remains a challenge for larger operations, necessitating ongoing improvements.