Red Hat Bugzilla – Bug 1124880
[fuse] java.io.FileNotFoundException (FNF) during time period with unrecovered disk errors
Last modified: 2015-03-05 07:36:29 EST
Description of problem: We experienced very infrequent FNF errors (3 occurrences in one month period) when running hadoop benchmarks: Wordcount, TestDFSIO read and Pi Estimator, on glusterfs. During the time period when the errors occurred we had a disk fail in our RAID6 array which the gluster volume resides on. Since we replaced the downed disk we have not been able to reproduce this problem. Version-Release number of selected component (if applicable): How reproducible: Cannot be reproduced at this time Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Upstream fix available at https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=154210ccb3a871e631bf39fdeb7a8731d98af87b
Note about the comment about disk replacement - the actual 'no such file or directory' (or 'File not Found' in Java lingo) seems not to be caused by the disk failure/replacement, but probably just impacted probability with timing/races because of delays.
A correction: 'File not Found' is a generic exception Java throws. It was found during debugging that the system errno was ESTALE and not ENOENT (Java just throws FNF exception even for ESTALE). Hopefully that clarifies the intention of the patch.
Patch(es) available on kernel-3.10.0-191.el7
Test failed on kernel-3.10.0-190.el7, got failed message: /mnt/one/file: Stale file handle Test passed on kernel3.10.0-220.el7
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHSA-2015-0290.html