Bug 1102369

Summary: Restarting Pulp after changing concurrency left stale workers
Product: [Retired] Pulp Reporter: Alex Krzos <akrzos>
Component: async/tasksAssignee: pulp-bugs
Status: CLOSED UPSTREAM QA Contact: pulp-qe-list
Severity: low Docs Contact:
Priority: low    
Version: 2.4.0CC: dgregor, mhrivnak, skarmark
Target Milestone: ---Keywords: Triaged
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2015-02-28 22:09:12 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:

Description Alex Krzos 2014-05-28 20:42:22 UTC 
Description of problem:
Change PULP_CONCURRENCY in /etc/default/pulp_workers prior to restarting pulp_resource_manager, pulp_workers, pulp_celerybeat has left several stale workers on my machine.

Version-Release number of selected component (if applicable):
mod_wsgi-3.4-1.pulp.el6sat.x86_64
pulp-puppet-tools-2.4.0-0.18.beta.el6sat.noarch
m2crypto-0.21.1.pulp-10.el6sat.x86_64
pulp-rpm-plugins-2.4.0-0.18.beta.el6sat.noarch
pulp-katello-plugins-0.3-1.el6sat.noarch
python-pulp-rpm-common-2.4.0-0.18.beta.el6sat.noarch
python-pulp-bindings-2.4.0-0.18.beta.el6sat.noarch
createrepo-0.9.9-21.2.pulp.el6sat.noarch
python-pulp-common-2.4.0-0.18.beta.el6sat.noarch
pulp-server-2.4.0-0.18.beta.el6sat.noarch
pulp-nodes-parent-2.4.0-0.18.beta.el6sat.noarch
pulp-nodes-common-2.4.0-0.18.beta.el6sat.noarch
pulp-selinux-2.4.0-0.18.beta.el6sat.noarch
python-isodate-0.5.0-1.pulp.el6sat.noarch
python-pulp-puppet-common-2.4.0-0.18.beta.el6sat.noarch
python-kombu-3.0.15-5.pulp.el6sat.noarch
pulp-puppet-plugins-2.4.0-0.18.beta.el6sat.noarch

How reproducible:
Always

Steps to Reproduce:
1. change PULP_CONCURRENCY to < available cores on machine in /etc/default/pulp_workers while pulp is running
2. service pulp_resource_manager restart; service pulp_workers restart; service pulp_celerybeat restart;
3. view processes to see if stale processes remain

Actual results:
Pulp only stops and restarts the number of workers as set to PULP_CONCURRENCY.

Expected results:
Pulp should stop all Pulp workers and start only the number specified in PULP_CONCURRENCY or if the value is commented out then the default.

Additional info:
Comment 1 Dennis Gregorovic 2015-02-10 14:30:58 UTC 
Just ran into this myself as well.
Comment 2 Brian Bouterse 2015-02-28 22:09:12 UTC 
Moved to https://pulp.plan.io/issues/440