Nice report. Thanks. I have a dev environment i build with vagrant every day. I even managed to bootstrap a puppetmaster in there for testing exported resources. Highly recommend using Vagrant because it forces you to fix those dependency errors you would normally not see on live servers that have been running for a while.
All my code gets the bugs tested out of it on base vms before it gets pushed into production. On 13 June 2014 02:38, Christopher Wood <christopher_w...@pobox.com> wrote: > Verdict: Went fine with the usual upgrade teething troubles. (Once I figured > those out I reverted to my pre-upgrade VM snapshot on the first upgraded host > and there was no hassle the second time.) I went from 3.4.3 to 3.6.2. > > > The procedure was to upgrade the following rpms on each host (daemon restarts > were included), reading logs as I went: > > for i in mcollective-puppet-agent hiera facter puppet mcollective; do mco > package $i update -F hostname=$SERVER; done > > Upgrading mcollective this way meant the last mco package never returned > since the daemon was restarted, I'm fine with that. > > Then this for the puppetmasters to ensure the passenger-based puppetmasters > were restarted: > > mco service httpd restart -F hostname=$SERVER; done > > > These three issues, unfixed, would have been blockers for me doing an upgrade: > > > 1) yumrepo mirrorlist vs baseurl (fixed in puppet 3.5.1) > > https://tickets.puppetlabs.com/browse/PUP-2155 (dupe) > https://tickets.puppetlabs.com/browse/PUP-2150 > > I was using mirrorlist, it was past time to use baseurl anyway. I never saw > an error since I was on 3.4.3. > > Solution: use baseurl. > > > 2) datacat module 0.4.2 incompatible with puppet >= 3.5 (fixed in datacat > 0.5.0) > > https://github.com/richardc/puppet-datacat/issues/12 > > I saw this sort of error on the puppetmaster: > > Jun 12 09:48:15 puppetmaster1dev puppet-master[4552]: undefined method > `module' for "production":String at > /etc/puppet/environments/production/modules/datacat/manifests/init.pp:59 on > node puppetmaster1dev.company.com > > Solution: upgrade the datacat module to 0.5.0 > > > 3) mcollective service/package trouble with puppet 3.5 (fixed in > mcollective-puppet-agent 1.7.1) > > https://tickets.puppetlabs.com/browse/MCOP-23 > > I saw the following errors when checking package status after the initial > aborted upgrade of mcollective only, not any mcollective plugins. > > a) First, this after a plain upgrade (but without restarting the puppet > agent): > > $ mco package mcollective status -F hostname=puppetmaster1dev > > * [ ==========================================================> ] 1 / 1 > > puppetmaster1dev.company.com: Could not autoload puppet/type/package: no > 'environments' in > {:root_environment=>#<Puppet::Node::Environment:0x7fc087f1d070 > @manifest="/tmp", @modulepath=[], @watching=true, @name=:"*root*", > @config_version=nil>, > :current_environment=>#<Puppet::Node::Environment:0x7fc087f1d070 > @manifest="/tmp", @modulepath=[], @watching=true, @name=:"*root*", > @config_version=nil>} at top of [[0, nil, nil]] > > Summary of Arch: > > No aggregate summary could be computed > > Summary of Ensure: > > No aggregate summary could be computed > > > Finished processing 1 / 1 hosts in 46.00 ms > > b) Second, after restarting both mcollective and puppet: > > $ mco package mcollective status -F hostname=puppetmaster1dev > > * [ ==========================================================> ] 1 / 1 > > puppetmaster1dev.company.com: undefined method `properties' for > nil:NilClass > > Summary of Arch: > > No aggregate summary could be computed > > Summary of Ensure: > > No aggregate summary could be computed > > > Finished processing 1 / 1 hosts in 31.84 ms > > > Solution: upgrade mcollective-puppet-agent before upgrading/restarting > mcollective and puppet > > > > Editorially, this was very relaxing and I should have saved it for Friday > afternoon. > > > On Thu, Jun 12, 2014 at 07:48:24AM -0400, Christopher Wood wrote: >> That's why I'm upgrading (otherwise 3.4.3 works for me), but I will have a >> hard time justifying any upgrade that leads to breakage. They're in >> profile::base for me too. >> >> Well, on with my testing! >> >> On Thu, Jun 12, 2014 at 10:36:07AM +1000, Pete Brown wrote: >> > I decided to put puppet,hiera and facter versions into my base profile >> > so upgrades are tested in my dev environment and then pushed out to >> > production when I am happy. >> > 3.6.2 is also a security release so I (and you very likely) want it >> > (and the relevant ssl fixes) pushed out asap. >> > I also just added a versioned ssl class to my profiles to make that even >> > easier. >> > >> > On 12 June 2014 06:40, Christopher Wood <christopher_w...@pobox.com> wrote: >> > > I only notice three issues where "Affects Version: 3.6.2". Have any of >> > > you upgraded to 3.6.2 yet? If so, where did you upgrade from and how's >> > > life now? >> > > >> > > https://tickets.puppetlabs.com/browse/PUP-2752?jql=affectedVersion%20%3D%20%223.6.2%22 >> > > >> > > (I'm probably doing this on our dev puppetmasters tomorrow, I will file >> > > tickets if there are any surprises.) >> > > >> > > -- >> > > You received this message because you are subscribed to the Google >> > > Groups "Puppet Users" group. >> > > To unsubscribe from this group and stop receiving emails from it, send >> > > an email to puppet-users+unsubscr...@googlegroups.com. >> > > To view this discussion on the web visit >> > > https://groups.google.com/d/msgid/puppet-users/20140611204001.GA20741%40iniquitous.heresiarch.ca. >> > > For more options, visit https://groups.google.com/d/optout. >> > >> > -- >> > You received this message because you are subscribed to the Google Groups >> > "Puppet Users" group. >> > To unsubscribe from this group and stop receiving emails from it, send an >> > email to puppet-users+unsubscr...@googlegroups.com. >> > To view this discussion on the web visit >> > https://groups.google.com/d/msgid/puppet-users/CAJ8DPF7PLa%2B3RNA8bO-E2aZcmhaMsTLZgZewE08sYAsJ7GQ8Fw%40mail.gmail.com. >> > For more options, visit https://groups.google.com/d/optout. >> >> -- >> You received this message because you are subscribed to the Google Groups >> "Puppet Users" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to puppet-users+unsubscr...@googlegroups.com. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/puppet-users/20140612114824.GB25610%40iniquitous.heresiarch.ca. >> For more options, visit https://groups.google.com/d/optout. > > -- > You received this message because you are subscribed to the Google Groups > "Puppet Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to puppet-users+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/puppet-users/20140612163813.GA26648%40iniquitous.heresiarch.ca. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "Puppet Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-users/CAJ8DPF4VGTqUpde4w7Bjf_U-Lb%2B%3DZM_hPwn3KBnx2WVrq-%3DiYA%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
