Pretty negligible per bot per request, but I’m not here to feed them. They also travel in packs, so the bandwidth does multiply. It also costs me money when I exceed my monthly bandwidth quota. I’ve blocked them for so long, I no longer have data I can tally to get an aggregate total (I only keep 90 days). SemrushBot alone, before I blocked it, was averaging about 15 GB a month. That one is fairly aggressive, though. Imagesift Bot, which pulls down any images it can find, would also use quite a bit, I imagine, if it were allowed.
With Lemmy, especially earlier versions, the queries were a lot more expensive, and bots hitting endpoints that triggered a heavy query (such as a post with a lot of comments) would put unwanted load on my DB server. That’s when I started blocking bot crawlers much more aggressively.
Static sites are a lot less impactful, and I usually allow those. I’ve got a different rule set for them which blocks the known AI scrapers but allows search indexers (though that distinction is slowly disappearing).
And by blocking search robots, do you stop being present in the search results or are you still present, but they do not show the content in question?
I block bots by default, and that prevents them from being indexed since they can’t be crawled at all. Searching “dubvee” (my instance name / url) in Google returns no relevant results. I’m okay with that, lol, but some people would be appalled.
However, I can search for things I’ve posted from my instance if they’ve federated to another instance that is crawled; the link will just be to the copy on that instance.
For the few static sites I run (mostly local business sites since they’d be on Facebook otherwise), I don’t enforce the bot blocking, and Google, etc are able to index them normally.
Pretty negligible per bot per request, but I’m not here to feed them. They also travel in packs, so the bandwidth does multiply. It also costs me money when I exceed my monthly bandwidth quota. I’ve blocked them for so long, I no longer have data I can tally to get an aggregate total (I only keep 90 days). SemrushBot alone, before I blocked it, was averaging about 15 GB a month. That one is fairly aggressive, though. Imagesift Bot, which pulls down any images it can find, would also use quite a bit, I imagine, if it were allowed.
With Lemmy, especially earlier versions, the queries were a lot more expensive, and bots hitting endpoints that triggered a heavy query (such as a post with a lot of comments) would put unwanted load on my DB server. That’s when I started blocking bot crawlers much more aggressively.
Static sites are a lot less impactful, and I usually allow those. I’ve got a different rule set for them which blocks the known AI scrapers but allows search indexers (though that distinction is slowly disappearing).
I block bots by default, and that prevents them from being indexed since they can’t be crawled at all. Searching “dubvee” (my instance name / url) in Google returns no relevant results. I’m okay with that, lol, but some people would be appalled.
However, I can search for things I’ve posted from my instance if they’ve federated to another instance that is crawled; the link will just be to the copy on that instance.
For the few static sites I run (mostly local business sites since they’d be on Facebook otherwise), I don’t enforce the bot blocking, and Google, etc are able to index them normally.
Thanks for the explanation and it was clear to me.