3 ms·
I couldn’t understand the scale of this “fediverse effect”. There is this quote: > ” making for a traffic amplification of 36704:1” Does it mean that posting
by soneca 2y ago
I couldn’t understand the scale of this “fediverse effect”.
There is this quote:
> ” making for a traffic amplification of 36704:1”
Does it mean that posting a link on Mastodon generates 36704 requests to that URL?
- skilled 2y agoSeveral thousand, yea. That many, unlikely in a single batch. Favouriting and boosting are actions that amplify how far a Mastodon post is going to spread on a “connected instances” spectrum. But for popular accounts (15k followers as in this case) will definitely spread the link to thousands of instances instantly and cause thousands of individual GET requests. It’s fun to tail access.log and see it happen in real time.
- SideburnsOfDoom 2y ago> Does it mean that posting a link on Mastodon generates 36704 requests to that URL? No, it's a ratio of bytes for traffic . They're counting size of 1 request to post to Mastodon "a single roughly ~3KB POST", vs the total size of content served from GETing the url in that post. IDK how valid this metric is, but that's what they're saying.
- kevincox 2y agoI don't see how the request size to mastodon is at all relevant to the resulting traffic.
- SideburnsOfDoom 2y agoIt's the language of "traffic amplification" as a network attack (1) as used in a DDOS. it's perhaps more relevant when comparing the total bytes in a number of IP Packets of various sizes. I'm not saying that this is the optimal framing, but that's what they were going for: talking about this, correctly or not, as a de facto DDOS. 1) https://www.microsoft.com/en-us/security/blog/2022/05/23/anatomy-of-ddos-amplification-attacks/ https://www.microsoft.com/en-us/security/blog/2022/05/23/ana...