tag:blogger.com,1999:blog-8263949408347549596.post7766651062057415493..comments2023-10-23T23:19:01.111+02:00Comments on Object-Oriented Software Development: HTTP/2, Keep-Alive and Safari browsers DDoSing your siteWiktor Zychlahttp://www.blogger.com/profile/04420514974154487039noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-8263949408347549596.post-12835034592562825642022-07-21T17:05:09.189+02:002022-07-21T17:05:09.189+02:00Thanks for this post. It got me pointed in the cor...Thanks for this post. It got me pointed in the correct direction for fixing an issue with Safari hammering repeated requests at our Node.js web apps. My <a href="https://stackoverflow.com/questions/71037910/safari-10-fails-to-load-https-with-node-js-iisnode-spams-requests" rel="nofollow">Stack Overflow post</a> has the details.Anonymoushttps://www.blogger.com/profile/11263032671896689286noreply@blogger.comtag:blogger.com,1999:blog-8263949408347549596.post-28733262965344823392022-05-30T15:15:17.024+02:002022-05-30T15:15:17.024+02:00Big thanks to your post here. We were plagued by a...Big thanks to your post here. We were plagued by a similar issue with certain versions of Safari and it took this post to get me started on the path of the HTTP/2 and header issue. In our case were were using IIS and Node.js (with IISNode) and the Node.js apps were sending back the "Connection" header over HTTP/1 but IIS was forwarding that back to the client over HTTP/2. I solved this by manually removing the "Connection" header. See https://stackoverflow.com/a/72434880/10229666 for more details.Anonymoushttps://www.blogger.com/profile/11263032671896689286noreply@blogger.comtag:blogger.com,1999:blog-8263949408347549596.post-57284074718924611242018-02-14T00:34:21.962+01:002018-02-14T00:34:21.962+01:00The bug is in the front server. The RFC is very cl...The bug is in the front server. The RFC is very clear on it:<br />"An endpoint MUST NOT generate an HTTP/2 message containing connection-specific header fields; any message containing connection-specific header fields MUST be treated as malformed (Section 8.1.2.6)."Ghttps://www.blogger.com/profile/04678396523967046572noreply@blogger.com