Vercel fast origin transfer (FOT) usage much higher than old bandwidth
Unanswered
Paper wasp posted this in #help-forum
Paper waspOP
Hi, I found something strange about new metrics related to Vercel's new pricing infra. Github discussion here with details/screenshots: https://github.com/orgs/vercel/discussions/6922
I'm seeing fast origin transfer (FOT) usage about 4x higher than fast data transfer (FDT) and also 4x higher than the old bandwidth usage. I'm quite confused how this is happening.
I use ISR with 12 hour revalidation on almost all my pages, and my app has consistent traffic. So my understanding is the ISR cache should be only affecting FDT usage, and so FDT should be much higher than FOT? I'm also confused how new FOT usage is 4x higher than old bandwidth usage, since I thought old bandwidth = FOT + FDT
Does anyone know what's going on here/is this expected? I feel like the FOT metric looks wrong.. It's hard to debug cause I can't breakdown FOT by paths
I'm seeing fast origin transfer (FOT) usage about 4x higher than fast data transfer (FDT) and also 4x higher than the old bandwidth usage. I'm quite confused how this is happening.
I use ISR with 12 hour revalidation on almost all my pages, and my app has consistent traffic. So my understanding is the ISR cache should be only affecting FDT usage, and so FDT should be much higher than FOT? I'm also confused how new FOT usage is 4x higher than old bandwidth usage, since I thought old bandwidth = FOT + FDT
Does anyone know what's going on here/is this expected? I feel like the FOT metric looks wrong.. It's hard to debug cause I can't breakdown FOT by paths