Fix memory leak with ruby/require/autoload_paths #3266
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
In ruby versions less than 2.5.0, it was found that if you have
Pathname objects in the $LOAD_PATH instead of regular strings, it messes
with ruby's internals and causes a memory leak.
In most ruby applications, this wouldn't be much of an issue, but there
are a significant number of deferred require statements that we do in
manageiq that cause this to leak overtime. Further more, this seems to
be an issue that presents itself even if the file has been loaded
previously (mostly a no-op for require).
Replication of this issue can be done using a simple ruby script:
require 'pathname'
require 'fileutils'
puts Process.pid
Dir.mkdir("foo") unless Dir.exists?("foo")
$LOAD_PATH.unshift(Pathname.new("foo"))
FileUtils.touch("empty.rb")
1500.times { 1500.times { require "empty" }; print "."; GC.start; }
By simply running the application with the Pathnames converted to
strings, this should be a proper and low cost workaround for us until it
is patched in ruby or we update manageiq to >= ruby 2.5.0.
|
@miq-bot add_label bug, performance |
3 tasks
kbrock
approved these changes
Jan 17, 2018
gtanzillo
approved these changes
Jan 17, 2018
|
Checked commit NickLaMuro@5fc7b05 with ruby 2.3.3, rubocop 0.52.0, haml-lint 0.20.0, and yamllint 1.10.0 |
|
@miq-bot add_labels fine/yes, gaprindashvili/yes |
simaishi
pushed a commit
that referenced
this pull request
Jan 19, 2018
Fix memory leak with ruby/require/autoload_paths (cherry picked from commit 62653f6) https://bugzilla.redhat.com/show_bug.cgi?id=1536658
|
Gaprindashvili backport details: |
simaishi
pushed a commit
that referenced
this pull request
Jan 19, 2018
Fix memory leak with ruby/require/autoload_paths (cherry picked from commit 62653f6) https://bugzilla.redhat.com/show_bug.cgi?id=1536672
|
Fine backport details: |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In ruby versions less than 2.5.0, it was found that if you have
Pathnameobjects in the$LOAD_PATHinstead of regular strings, it messes with ruby's internals and causes a memory leak.In most ruby applications, this wouldn't be much of an issue, but there are a significant number of deferred require statements that we do in
manageiqthat cause this to leak overtime. Further more, this seems to be an issue that presents itself even if the file has been loaded previously (mostly a no-op forrequire).Replication of this issue can be done using a simple ruby script:
By simply running the application with the Pathnames converted to strings, this should be a proper and low cost workaround for us until it is patched in ruby or we update
manageiqto >= ruby 2.5.0.Links
Fixes https://bugzilla.redhat.com/show_bug.cgi?id=1535720
Steps for Testing/QA
Being tested in ManageIQ/manageiq#16837