+++ This bug was initially created as a clone of Bug #1379285 +++ Description of problem: From the code inspection, have observed that there are fd ref leaks for few fd-based gfapi fops. 'glfs_resolve_fd' returns a fd (either existing or migrated) with an extra ref taken. This needs to be unref'ed at the end of the operation. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: --- Additional comment from Worker Ant on 2016-09-26 05:35:39 EDT --- REVIEW: http://review.gluster.org/15573 (gfapi: Fix few fd ref leaks) posted (#1) for review on master by soumya k (skoduri) --- Additional comment from Worker Ant on 2016-09-27 02:43:38 EDT --- COMMIT: http://review.gluster.org/15573 committed in master by Niels de Vos (ndevos) ------ commit d27cffab670858d7812bbb458a0833303d009b3b Author: Soumya Koduri <skoduri> Date: Mon Sep 26 14:54:27 2016 +0530 gfapi: Fix few fd ref leaks From the code inspection, have observed that there are fd ref leaks for few fd-based gfapi fops. 'glfs_resolve_fd' returns a fd (either existing or migrated) with an extra ref taken. This needs to be unref'ed at the end of the operation. Change-Id: Id63394e3e7deafb0c8b06444f2ae847248b126db BUG: 1379285 Signed-off-by: Soumya Koduri <skoduri> Reviewed-on: http://review.gluster.org/15573 NetBSD-regression: NetBSD Build System <jenkins.org> CentOS-regression: Gluster Build System <jenkins.org> Reviewed-by: Niels de Vos <ndevos> Smoke: Gluster Build System <jenkins.org>
Upstream mainline : http://review.gluster.org/15573 Upstream 3.8 : http://review.gluster.org/15585 downstream : https://code.engineering.redhat.com/gerrit/86107
Ran Basic sanity on ganesha and don't see any issues with fd leak. Will have inputs from Samba and Qemu (users of libgfapi), if basic sanity works fine there , will move the BZ to verified.
Samba ----- Ran basic sanity including mount, unmount, IOs, share connect disconnects and did not see any issues with fd leak.
Tested RHGS 3.2.0 interim build ( glusterfs-3.8.4-5.el7rhgs ) with RHEL 7.3 with following components : qemu-img-rhev-2.6.0-27.el7.x86_64 qemu-kvm-rhev-2.6.0-27.el7.x86_64 libvirt-2.0.0-10.el7.x86_64 I could able to create VMs and ran I/O inside the VMs. All works good. Note: I haven't tested for the memory footprint or fd leaks with gfapi
As there are no regressions seen on nfs-ganesha, samba and Qemu because of this patch, moving the BZ to verified. nfs-ganesha-2.4.1-1.el7rhgs.x86_64 nfs-ganesha-gluster-2.4.1-1.el7rhgs.x86_64 glusterfs-ganesha-3.8.4-5.el7rhgs.x86_64 glusterfs-api-3.8.4-5.el7rhgs.x86_64
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-2017-0486.html