On Fri, Feb 14, 2014 at 08:55:11AM +0000, Wuqixuan wrote: > Hi Trond, Bfields, All, > > We are useing nfsv3 in 2.6.34.13 kernel for commucial use. For some reasons(such as security), has requirement to upgrade to nfsv4(still on 2.6.34.x kernel). We did two things: > > 1. I passed through the git log of kernel.org of nfs, found after 2010(2.6.34 is released on May 2010, and no update on 2.6.34.x for nfs/nfsd), just for sampling, nfsd/nfs4proc.c has 460 commits while nfs/nfs3proc.c has only 38 commits. > 2. Just enable nfsv4 on 2.6.34.13, got problem(EIO) on just a simple open. Found it's because of one commit(https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=a2c0b9e291208f65221a0ad8a0c80a377707d480), simply remove this commit, now can simply use nfsv4 on 2.6.34.13. > > So we are worried about the stability of nfsv4 on 2.6.34.x. > If any one has some idea or information about: > > 1. Whether nfsv4 (including nfs/nfs4/sunrpc) of 2.6.34.x is stable enough for commercial use or not. > 2. If nfsv4 is not stable enough on 2.6.34.x, which version can we use , 3.10 ? or 3.12 ? > 3. Is there any mature commercial application on nfsv4 on 2.6.34.x or before? Can tell some famous company or app name ? I don't know anything about the specific bug that you found. Based on https://lwn.net/Articles/585416/ 2.6.34 appears to be EOL'd as of 2.6.34.15, so it would be up to you to support it past that. http://www.kernel.org/ lists which stable kernel branches are still being updated. I'm not sure where you find out how long those branches are expected to be maintained. Someone who needs guaranteed support periods for "commercial use" usually gets a contract with someone like my employer. (And indeed the enterprise distros do support NFSv4 on kernels that branched off before 2.6.34). --b. -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
