Description of problem: On AH 7.4.0, we can't query rpm packages by command "rpm -qa" on the host, this lead to installer failed when trying to install NTP mistakenly on AH. Version-Release number of selected component (if applicable): openshift-ansible-3.6.121-1.git.0.ed0b72c.el7.noarch.rpm # atomic host status State: idle Deployments: ● rhel-atomic-host:rhel-atomic-host/7/x86_64/standard Version: 7.4.0 (2017-06-16 21:48:31) Commit: 0b62181fb351d26ebb794438174c8dbc2ea5cc4c04c589b7be82e15695a2bbbb `rpm -qa` returned nothing # rpm -qa How reproducible: always Steps to Reproduce: 1. Trigger installation against AH 7.4 Actual results: Installer mistakenly checked that "openshift.clock.chrony_installed" was false due to command "rpm -q" is not useful any more on AH 7.4: # rpm -q chrony package chrony is not installed Actually the service is installed. Then installer tried to install NTP package on AH, but that's not available on AH. TASK [openshift_clock : Install ntp package] *********************************** Friday 23 June 2017 08:45:48 +0000 (0:00:05.441) 0:12:25.215 *********** fatal: [openshift-133.lab.sjc.redhat.com]: FAILED! => { "changed": false, "failed": true } MSG: Could not find a module for unknown. fatal: [openshift-151.lab.sjc.redhat.com]: FAILED! => { "changed": false, "failed": true } MSG: Could not find a module for unknown. fatal: [openshift-124.lab.sjc.redhat.com]: FAILED! => { "changed": false, "failed": true } MSG: Could not find a module for unknown. Expected results: As Atomic Host is immutable, we don't have to check whether the *rpm* packages are installed or not. Should skip the rpm checks on AH. Additional info:
*** This bug has been marked as a duplicate of bug 1462979 ***
(In reply to Scott Dodson from comment #2) > > *** This bug has been marked as a duplicate of bug 1462979 *** Not sure how this bug is duplicate of bug 1462979 because that is RHEL rpm specific. We are getting the same error while installing OpenShift 3.6 on RHEL Atomic, is there any workaround for this? TASK [openshift_clock : Set clock facts] *************************************** changed: [lnyce095xsdi008.sdi.corp.example.com] TASK [openshift_clock : Install ntp package] *********************************** fatal: [lnyce095xsdi008.sdi.corp.example.com]: FAILED! => { "changed": false, "failed": true } MSG: Could not find a module for unknown. NO MORE HOSTS LEFT ************************************************************* to retry, use: --limit @/root/openshift/infra-maint/roles/pre_upgrade/ose-3.6/usr/share/ansible/openshift-ansible/playbooks/byo/openshift-master/scaleup.retry
Shah, Which commit of atomic host are you at? `atomic host status` please. The only way that task would've executed is if chrony weren't installed and I've just looked at 7.4.0 GA and 7.4.1 both have chrony. 7.4.0 = 846fb0e18e65bd9a62fc9d952627413c6467c33c2d726449a1d7ad7690bbb93a 7.4.1 = e83c16780259c5272684221e2a6007300d94bbfdc5432f9ab6025300f447145b
(In reply to Scott Dodson from comment #4) > Shah, > > Which commit of atomic host are you at? `atomic host status` please. The > only way that task would've executed is if chrony weren't installed and I've > just looked at 7.4.0 GA and 7.4.1 both have chrony. > > 7.4.0 = 846fb0e18e65bd9a62fc9d952627413c6467c33c2d726449a1d7ad7690bbb93a > 7.4.1 = e83c16780259c5272684221e2a6007300d94bbfdc5432f9ab6025300f447145b I think I have to update the AH: -bash-4.2# atomic host status State: idle Deployments: ● bac-atomic-host:rhel-atomic-host/7/x86_64/standard Version: 7.3.3 (2017-02-27 16:31:38) Commit: bfc591ba1a4395c6b8e54d34964b05df4a61e0d82d20cc1a2fd817855c7e2da5 OSName: rhel-atomic-host rhel-atomic-host:rhel-atomic-host/7/x86_64/standard Version: 7.3.2 (2017-01-13 22:00:41) Commit: 96826a0d917d7ff10f9fd0289581649f2ffbddd76f3b80efd3d95cc11915cacb OSName: rhel-atomic-host
Is your RPMDB corrupt? what do you get from `rpm -q chrony` ? Our QE teams would've tested 7.3.3 at some point in the past and this is a pretty obvious failure that I'm sure they would've encountered if RHEL AH 7.3.3 lacked chrony, in fact this bug was opened by our QE team who ran into such a problem in an early 7.4.0 build.
You are right, chrony is not installed. We are trying to upgrade the host 7.4.1 -bash-4.2# rpm -q chrony package chrony is not installed Thanks for the notes and help.