Description of problem: The command was working fine and suddenly it stopped launching console, execution fails with error : [root@ieabvlxrcfmel01 ~]# appliance_console /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/spec_set.rb:87:in `block in materialize': Could not find nokogiri-1.6.8 in any of the sources (Bundler::GemNotFound) from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/spec_set.rb:81:in `map!' from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/spec_set.rb:81:in `materialize' from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/definition.rb:159:in `specs' from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/definition.rb:218:in `specs_for' from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/definition.rb:207:in `requested_specs' from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/runtime.rb:109:in `block in definition_method' from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler/runtime.rb:21:in `setup' from /opt/rh/cfme-gemset/gems/bundler-1.15.0/lib/bundler.rb:101:in `setup' from /bin/appliance_console:4:in `<main>' [root@ieabvlxrcfmel01 ~]# uptime Build : 5.8.0.17-20170525183055_6317a22 Nokogiri : rh-ruby23-rubygem-nokogiri-1.6.8-1.el7cf.x86_64 cfme-gemset : cfme-gemset-5.8.0.17-1.el7cf.x86_64 gem : nokogiri (1.6.8) rpm -V rh-ruby23-rubygem-nokogiri-1.6.8-1.el7cf.x86_64 no modifiction noticed.
> The command was working fine and suddenly it stopped launching console Were there any package updates? I can't tell from the support case or the sos logs if there was a cfme upgrade performed, if any packages were updated, or nothing like this. Was bundle install run? Note, you can't run bundle install as this will invalidate the dependency tree for cfme and found in the Gemfile.lock. Let's see if the bundle install was run: rpm -V cfme-gemset If there is a change to the Gemfile.lock or anything else in this package, we need to reinstall it: yum reinstall cfme-gemset # if there are any changes found in the first command Next, you can try looking for any changes in any of the cfme rpms: rpm -qa |grep cfme | xargs rpm -V $1 In my environment, the only ones I see changed is this: S.5....T. /var/www/miq/vmdb/tmp/cache/sti_loader.yml I'm not sure what else to check. Please let me know what you find.
The Gemfile.lock was modified by something and the solution they're going to use is yum reinstall cfme-gemset to get it back to the way it was.
Created attachment 1454331 [details] rpm_package_nonworking_appliance
Created attachment 1454332 [details] rpm_package_working_appliance
Created attachment 1454333 [details] rpm_verification_nonworking_appliance
Created attachment 1454334 [details] rpm_verification_working_appliance
Prasad, it looks like the Gemfile.lock is wrong on the non-working appliance. We use nokogiri 1.8.1 on cfme 5.7.4.2. The error you're reporting is looking for version 1.6.8. Please check the Gemfile.lock in /var/www/miq/vmdb/Gemfile.lock. Note, this is a symlink to the cfme-gemset's Gemfile.lock There should be a line that says: nokogiri (= 1.8.1) If it says 1.6.8, you need to yum reinstall cfme-gemset to fix it. If it says 1.8.1, maybe try running appliance_console from /var/www/miq/vmdb and report if that works.
Note, in our private chat, you mentioned the error: Could not find nokogiri-1.8.1 in any of the sources. That is a completely different problem from the reported one for 1.6.8. Please double check what the error is. The Gemfile.lock should say 1.8.1 as per comment 10. You should have nokogiri installed, which is confirmed by the package list you provided for both the working non-working appliances: 10:26:39 ~/Downloads/issues/bz1571827 (2.4.4) + grep nokogiri rpm_package_* rpm_package_nonworking_appliance.txt:rh-ruby23-rubygem-nokogiri-1.8.1-2.el7cf.x86_64 rpm_package_working_appliance.txt:rh-ruby23-rubygem-nokogiri-1.8.1-2.el7cf.x86_64