Description of problem: ======================= While accessing 256 snapshots under .snaps from fuse and NFS , some snapshots intermittently gave ENOENT from fuse and 'Invalid Argument' error from NFS Version-Release number of selected component (if applicable): ============================================================= glusterfs-3.6.0.42.1-1.el6rhs.x86_64 How reproducible: ================ Often Steps to Reproduce: =================== 1.Create a 6x2 dist rep volume and start it 2.Fuse and NFS mount the volume and create heavy IO 3.Create 256 snapshots in loop and activate them from another node 4.From the client access .snaps and cd to every snap in .snaps directory from the / of the volume and from random sub-directories 5.Some snaps intermittently gave 'ENOENT' from fuse and 'Invalid Argument' from NFS mount . After clearing the cache , snapshots were successfully accessed under .snaps Actual results: ============== Some snapshots intermittently gave ENOENT from fuse and 'Invalid Argument' error from NFS Expected results: ================ There should be no errors seen while trying to access snapshots under .snaps Additional info:
Added doc text: Cause: When a large number of snapshots are activated and are accessed via USS, the inode limit for the mounted filesystem may get exhausted, causing deletion of inode entries from the inode table. Accessing .snaps when this happens may result in ENOENT errors in fuse, and "Invalid Argument" errors in NFS. Workaround: Clear the kernel VFS cache by executing the following command: # echo 3 >/proc/sys/vm/drop_caches
sosreports : ---------- http://rhsqe-repo.lab.eng.blr.redhat.com/bugs_necessary_info/snapshots/1191033/
The information is added as a Known Issue in the "Release Notes": http://docbuilder.usersys.redhat.com/22629/#Red_Hat_Storage Do review the same and share your comments if any.
Looks great. Thanks Bhavana.
This is fixed in the doc as verified in comment 5. Hence closing this bug.