Local file downloads reflect the current link from a specific device to the target data center; the multi-location delay table displays the round-trip time for each of the multiple measurement points. Both results are only suitable for archiving if the entry point truly belongs to the target data center.
When comparing candidate data centers, try to keep file size, access provider, and test period the same. If you want to serve overseas visitors, add target market test points to avoid using a single domestic sample to replace the real user distribution.
Located on the East Coast of the United States, historical line measurements show that direct connection and lower latency are two indicators that need to be considered separately.
Download speed, round-trip latency, packet loss, and path length each answer different questions, making them suitable to be included in the same measurement record.