[LeonM]

> google bots scour the web relentlessly and hammer sites almost to death

I have been hosting sites and online services for a long time now and never had this problem, or heard of this issue ever before.

If your site can't even handle a crawler, you need to seriously question your hosting provider, or your architecture.

[sbarre]

Perhaps stop and reconsider such a dismissive opinion given that "you've never had this issue before" then? Or go read up a bit more on how crawlers work in 2023.

If your site is very popular and the content changes frequently, you can find yourself getting crawled a higher frequency than you might want, particularly since Google can crawl your site at a high rate of concurrency, hitting many pages at once, which might not be great for your backend services if you're not used to that level of simultaneous traffic.

"Hammered to death" is probably hyperbole but I have worked with several clients who had to use Google's Search Console tooling[0] to rate-limit how often Googlebot crawled their site because it was indeed too much.

0: https://developers.google.com/search/docs/crawling-indexing/...

[holoduke]

I have a website thats get crawled at least 50 times per second. Is that a real deal? No not really. The site is probably doing 10.000 requests per second. I mean a popular site is indexed a lot. Your webserver should be designed for it. What tech are you using if I may ask?

[sbarre]

My specific case doesn't really matter (and my examples are from some years ago and of smaller clients, not my own setup).

My point was that people provision capacity ideally based on observed or expected traffic, and that crawlers can, and do, show up and exceed that capacity sometimes, having a negative effect on your customers' experience.

But you are correct that it's absolutely manageable. And telling crawlers to slow the F down is one of the tools you can use to manage it. :-)

[pas]

if your site is popular and you have a problem with crawlers use robots.txt (in particular the Crawl-delay stanza)

also for less friendly crawlers a rate limiter is needed anyway :(

(of course the existence of such tools doesn't give carte blanche to any crawler to overload sites ... but let's say they implement some sensing, based on response times, that means a significant load is probably needed to increase response times, which definitely can raise some eyebrows, and with autoscaling can cost a lot of money to site operators)

[piva00]

I worked at a company back in 2005-2010 where we had a massive problem with Googlebot crawlers hammering our servers, stuff like 10-100x the organic traffic.

That's pre-cloud ubiquity so scaling up meant buying servers, installing them on a data center, and paying rent for the racks. It was a fucking nightmare to deal with.