Bug 1079940
Summary: | GS should free memory allocated for application after it exists | ||
---|---|---|---|
Product: | Red Hat Enterprise Linux 7 | Reporter: | Vladimir Benes <vbenes> |
Component: | gnome-shell | Assignee: | Florian Müllner <fmuellner> |
Status: | CLOSED DUPLICATE | QA Contact: | Desktop QE <desktop-qa-list> |
Severity: | high | Docs Contact: | |
Priority: | medium | ||
Version: | 7.0 | CC: | afox, anton4linux, brendan.shephard, cww, elliot.li.tech, fmuellner, gonzo, jkoten, lmiksik, mark.crossland, mclasen, mdomonko, rstrode, tpelka, vhumpa, zagar |
Target Milestone: | rc | ||
Target Release: | 7.0 | ||
Hardware: | x86_64 | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | If docs needed, set a value | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2019-01-29 16:13:45 UTC | Type: | Bug |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: | |||
Bug Depends On: | |||
Bug Blocks: | 1649362, 1656436 |
Description
Vladimir Benes
2014-03-24 11:02:49 UTC
Can confirm on Fedora 21 (Gnome-Shell 3.14) Guys, looks like what you're describing here could be related to this: https://gitlab.gnome.org/GNOME/gnome-shell/issues/64 We have recently done and run a gnome-shell/gnome longetivity test that ran for 24 hours on fresh RHEL-7.5 And we indeed are seeing the issue that is heavily discussed on https://gitlab.gnome.org/GNOME/gnome-shell/issues/64. What the test does it runs apps in sequence via the Alt-F2 and then closes them using he App menu -> quit (thus is draws the app menu every time). gnome-shell process grows from early 1.2 GB to final 4.8 GB over this 24h period. Screenshot from kibana with the process monitor data: https://screenshots.firefox.com/Zc95rC6eaG1XwDaD/str-ha-prod-2.rhev-ci-vms.eng.rdu2.redhat.com RH internal users can access the reports here https://url.corp.redhat.com/b027f7d and here https://url.corp.redhat.com/6c27a14 (kibana/kibana). The system used is a openstack VM instance and the system with this session is kept running, so feel free to ping me for access. version: gnome-shell-3.26.2-5.el7.x86_64 (7.5 gold) I believe that if the upstream hunt for these leaks comes to a resolution, that might be a candidate for backporting to 7.5. It now appears that this issue has been resolved upstream. Links to the issues are: https://feaneron.com/2018/04/20/the-infamous-gnome-shell-memory-leak/ There is a useful summary of the issue by a developer (Philip Chimento) at: https://gitlab.gnome.org/GNOME/gnome-shell/issues/64#note_296648 Georges Stavracas explains the "Tardy Sweep" problem in his blog post. https://feaneron.com/2018/04/20/the-infamous-gnome-shell-memory-leak/ This is the merge request with Georges Stavracas fix for milestone "GNOME 3.28 bugfixes" https://gitlab.gnome.org/GNOME/gjs/merge_requests/114 I would ask that we backport this to both 7.5 and 7.6, as I have TAM customers who have large numbers of RHEL workstations users who see this problem daily. Probably dup of https://bugzilla.redhat.com/show_bug.cgi?id=1546302 |