Hello Developers, Firstly, thanks for your ongoing work on OpenAFS!
I'm Mark, a researcher currently evaluating it for a project I'm working on. I love several features like the online backup volumes and transparent volume moves -- very impressive and rather attractive to admins. However while OpenAFS seems to have some great featuers, the performance of it seems to be sorely lacking compared to peer products (like NFS, Ceph, Gluster). It's becoming harder to take performance of 100MB/s [+/-, usually -] seriously in the days of commodity 10Gb networking. I noticed that several patches on gerrit and old workshop presentations (sendmmsg, recvmmsg, maybe tcp-oob) exist which if implemented would help OpenAFS gain both performance and relevance in the network filesystem space. But it seems no traction has taken place to implement any of these despite being submitted years(!?) ago. I realize that it's important to maintain compatibility with the newest operating systems, but unless you put more effort towards keeping up with modern performance, more sites will drop OpenAFS and fewer new projects like mine will consider it. So, can some of the performance patches in gerrit please be merged into master for a production release of OpenAFS within the short term. Thanks for reading,Mark Sent with [Proton Mail](https://proton.me/) secure email.
