Hi Ced, no, it can not. However, spec defines two attributes, maxread and maxwrite, that can indicate client preferred IO request sizes: https://datatracker.ietf.org/doc/html/rfc5661#section-5.8.2.20 In pNFS case, an additional layout-specific option, that can override those values. Best regards, Tigran. ----- Original Message ----- > From: "Cedric Blancher" <cedric.blancher@xxxxxxxxx> > To: "Linux NFS Mailing List" <linux-nfs@xxxxxxxxxxxxxxx> > Sent: Wednesday, 3 April, 2024 19:59:16 > Subject: Linux NFSv4.1 client with Linux NFSv4.1 server, getting block size of exported filesystem? > Good evening! > > Assuming a Linux 6.6 NFSv4.1 client and a Linux 6.6 NFSv4.1 server, > can the NFSv4.1 client obtain the exact block size of the exported > filesystem (e.g. ReiserFS, btrfs, ext4, ...) via fstatat()/stat()? > > Ced > -- > Cedric Blancher <cedric.blancher@xxxxxxxxx> > [https://plus.google.com/u/0/+CedricBlancher/] > Institute Pasteur
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature
