This morning in logs I noticed a user got rate limited by our throttler 3 times within 1 second (429 too many requests error) at 3:30:11 AM EST: Mozilla/5.0 (Linux; Android 6.0; GCE x86 phone Build/MRA59G.MZC35) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2679.0 Safari/537.36
at IP 104.196.83.207
. I've only seen Google on this IP once before, so I'm unsure what it's purpose is.
Upon looking it up it seems it's a Googlebot indeed. Noted below by Analog, its coming from Google Cloud. The last paths it was trying before being limited for 30 seconds were images. We have never had problems with Google speeding the site until last night. Our throttler is set to trigger a 30 sec 429 error if a user hits 10 requests in a 5 second window (session var based). 3 logs imply they hit this limit 3 times within 1 second (30 total requests per sec).
Right now, the Search Console setting says it's on the "let Google decide" -- clicking the radio to change rate puts it square in the middle at "1 request every 10 seconds". On its maximum setting it allows 2 requests a second (ie throttle zone).
I will adjust the crawl rate manually to 1 every 10 seconds, but in the meantime why is this bot crawling so fast, and why is it on android mobile crawling only images?