Description of problem: When customer patches around 60 servers simultaneously from Satellite, many of them get to the "Uploading Tracer Profile" step around the same time, and some clients have delays of over 15 minutes. It is also possible to reproduce this by running a remote execution "katello-tracer-upload" command on multiple hosts. Version-Release number of selected component (if applicable): Satellite 6.3.5 How reproducible: Consistently Steps to Reproduce: 1. Have around 160 Satellite clients with katello-host-tools-tracer installed. 2. Create a Remote Execution job to run on the 160 clients and run the command "time katello-tracer-upload", with a 15 minute time out. Actual results: In customers environment, 3 of the servers timed out after 15 minutes, and 2 servers ran for over 5 minutes. Expected results: Servers not taking over 15 minutes to run katello-host-tools-tracer Additional info: Customer ran the same Remote execution job, but set the concurrency level to 1, so that it only ran on one server at a time, and did not have any failures or delays. See also: https://community.theforeman.org/t/katello-tracer-troubleshoot/8676
Created redmine issue https://projects.theforeman.org/issues/29397 from this bug
Upstream bug assigned to jturel
Moving this bug to POST for triage into Satellite 6 since the upstream issue https://projects.theforeman.org/issues/29397 has been resolved.
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 (Important: Satellite 6.8 release), 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://access.redhat.com/errata/RHSA-2020:4366