On Mon, 2 Sept 2024 at 12:50, Bernd Schubert <bernd.schubert@xxxxxxxxxxx> wrote: > In case of distributed servers, it can easily happen that one server has > an issue, while other servers still process requests. Especially when > these are just requests that read/getattr/etc and do not write, i.e. > accessing the stuck server is not needed by other servers. So in my > opinion not so unlikely. Although for such cases not difficult to > timeout within the fuse server. Exactly. Normally the kernel should not need to time out fuse requests, and it might be actively detrimental to do so. The main case this wants to solve is a deadlocked server due to programming error, AFAICS. And this would only work in environments where requests are guaranteed to complete within some time period. So if the server needs to handle request timeouts, then it should *not* rely on the kernel timeout. The kernel timeout should be a safeguard against broken or malicious servers, not an aid to implement request timeouts. Thanks, Miklos
