In some very specific circumstances, you might experience 429 errors when trying to resize many images simultaneously.
Cloudimage's goal is to provide you with the fastest and most reliable image resizing service and CDN on the market. Our resizing infrastructure is scaled to support 20x the highest load we experience on over the day. This allows us to guarantee SLA of 99% and let you forget about us. However in some rare circumstances, like for example network latencies between our resizing infrastructure and your image origin (for example AWS => Azure), we get CPU spikes which are 100x or even 200x the highest load we take into account.
To protect you from slow resizes and avoid downtime, we have an automated remediation system, that will prioritize tokens and throttle them as well as the tokens causing the high CPU spikes. In such a rare scenario, 429 errors are returned before a resize is even attempted to protect the overall architecture and guarantee performance.
Tokens running on FREE plans are usually prioritized lower, hence we recommend switching to a paid plan if you want to benefit from our standard SLAs.