Red Hat Bugzilla – Bug 1387658
Planning of remote execution task on large number of hosts (reported with 500 hosts) takes a lot of time, potentially leading to stuck task
Last modified: 2017-06-27 09:06:27 EDT
Description of problem: When trying to run a command on all the hosts registered to the Satellite by going to run job and putting * in the search field on the Satellite WebUI, then the task is in pending state for a long time and when Cu go over Tasks and click on the task created he get the following error: RuntimeError plan_self has to be invoked before being able to reference the output Whereas when running the command is run on a single host, then it is executed normally. Version-Release number of selected component (if applicable): V 6.2 How reproducible: Always Steps to Reproduce: 1.Login on Satellite Server --> Monitor --> Jobs --> Run job 2.In the search query, putting * and in the commands section, a command to be run on all the hosts 3.Where as if in the search section, a few hosts are defined then the command on these hosts is executable well. Actual results: I am not able to run a command on all the hosts at a time by entering * in the search section. Expected results: We should be able to run a command on all the hosts at a time by putting * in the search query section. Additional info:It works with few number of hosts by specifying the hosts but fails with * (large number of hosts)
The original error produced when looking at a task in pending state has been addressed in https://bugzilla.redhat.com/show_bug.cgi?id=1364767 is already shipped as part of errata https://access.redhat.com/errata/RHBA-2016:2940 (sat 6.2.5). Is seems the issue might be with the effectivity of the planning when working with larger number of hosts.
Created redmine issue http://projects.theforeman.org/issues/17921 from this bug
Upstream bug assigned to sshtein@redhat.com
There is also https://bugzilla.redhat.com/show_bug.cgi?id=1416542. We need to do more testing/tuning on large scale infrastructure to be able to say, if this is the only change needed, or additional updates are required.
Moving this bug to POST for triage into Satellite 6 since the upstream issue http://projects.theforeman.org/issues/17921 has been resolved.
Verified in Satellite 6.2.10 Snap 1 Kicked off a job against 700 hosts. The jobs started executing almost immediately. All jobs completed in under 8 minutes. See attached screenshot.
Created attachment 1282303 [details] verification screenshot
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://access.redhat.com/errata/RHBA-2017:1553