diff options
71 files changed, 1454 insertions, 277 deletions
diff --git a/.tito/packages/openshift-ansible b/.tito/packages/openshift-ansible index 271e86d20..4ec54c846 100644 --- a/.tito/packages/openshift-ansible +++ b/.tito/packages/openshift-ansible @@ -1 +1 @@ -3.0.28-1 ./ +3.0.35-1 ./ diff --git a/README_AWS.md b/README_AWS.md index f8ecaec49..c605de43d 100644 --- a/README_AWS.md +++ b/README_AWS.md @@ -51,7 +51,7 @@ to setup a private key file to allow ansible to connect to the created hosts. To do so, add the the following entry to your $HOME/.ssh/config file and make it point to the private key file which allows you to login on AWS. ``` Host *.compute-1.amazonaws.com - PrivateKey $HOME/.ssh/my_private_key.pem + IdentityFile $HOME/.ssh/my_private_key.pem ``` Alternatively, you can configure your ssh-agent to hold the credentials to connect to your AWS instances. @@ -16,6 +16,7 @@ Options: -c CLUSTER, --cluster CLUSTER which cluster to use -e ENV, --env ENV which environment to use + --v3 When working with v3 environments. v2 by default -t HOST_TYPE, --host-type HOST_TYPE which host type to use --list-host-types list all of the host types @@ -72,6 +72,8 @@ class Ossh(object): parser.add_argument('-o', '--ssh_opts', action='store', help='options to pass to SSH.\n \ "-oForwardX11=yes,TCPKeepAlive=yes"') + parser.add_argument('-A', default=False, action="store_true", + help='Forward authentication agent') parser.add_argument('host', nargs='?', default='') self.args = parser.parse_args() @@ -177,6 +179,9 @@ class Ossh(object): if self.user: ssh_args.append('-l%s' % self.user) + if self.args.A: + ssh_args.append('-A') + if self.args.verbose: ssh_args.append('-vvv') diff --git a/docs/best_practices_guide.adoc b/docs/best_practices_guide.adoc index 6b744333c..267aa850d 100644 --- a/docs/best_practices_guide.adoc +++ b/docs/best_practices_guide.adoc @@ -13,9 +13,12 @@ This guide complies with https://www.ietf.org/rfc/rfc2119.txt[RFC2119]. == Pull Requests + + +[[All-pull-requests-MUST-pass-the-build-bot-before-they-are-merged]] [cols="2v,v"] |=== -| **Rule** +| <<All-pull-requests-MUST-pass-the-build-bot-before-they-are-merged, Rule>> | All pull requests MUST pass the build bot *before* they are merged. |=== @@ -30,9 +33,10 @@ The tooling is flexible enough that exceptions can be made so that the tool the === Python Source Files ''' +[[Python-source-files-MUST-contain-the-following-vim-mode-line]] [cols="2v,v"] |=== -| **Rule** +| <<Python-source-files-MUST-contain-the-following-vim-mode-line, Rule>> | Python source files MUST contain the following vim mode line. |=== @@ -48,9 +52,10 @@ If mode lines for other editors are needed, please open a GitHub issue. === Method Signatures ''' +[[When-adding-a-new-paramemter-to-an-existing-method-a-default-value-SHOULD-be-used]] [cols="2v,v"] |=== -| **Rule** +| <<When-adding-a-new-paramemter-to-an-existing-method-a-default-value-SHOULD-be-used, Rule>> | When adding a new paramemter to an existing method, a default value SHOULD be used |=== The purpose of this rule is to make it so that method signatures are backwards compatible. @@ -74,18 +79,20 @@ def add_person(first_name, last_name, age=None): http://www.pylint.org/[PyLint] is used in an attempt to keep the python code as clean and as managable as possible. The build bot runs each pull request through PyLint and any warnings or errors cause the build bot to fail the pull request. ''' +[[PyLint-rules-MUST-NOT-be-disabled-on-a-whole-file]] [cols="2v,v"] |=== -| **Rule** +| <<PyLint-rules-MUST-NOT-be-disabled-on-a-whole-file, Rule>> | PyLint rules MUST NOT be disabled on a whole file. |=== Instead, http://docs.pylint.org/faq.html#is-it-possible-to-locally-disable-a-particular-message[disable the PyLint check on the line where PyLint is complaining]. ''' +[[PyLint-rules-MUST-NOT-be-disabled-unless-they-meet-one-of-the-following-exceptions]] [cols="2v,v"] |=== -| **Rule** +| <<PyLint-rules-MUST-NOT-be-disabled-unless-they-meet-one-of-the-following-exceptions, Rule>> | PyLint rules MUST NOT be disabled unless they meet one of the following exceptions |=== @@ -95,9 +102,10 @@ Instead, http://docs.pylint.org/faq.html#is-it-possible-to-locally-disable-a-par 1. When PyLint fails, but the code makes more sense the way it is formatted (stylistic exception). For this exception, the description of the PyLint disable MUST state why the code is more clear, AND the person reviewing the PR will decide if they agree or not. The reviewer may reject the PR if they disagree with the reason for the disable. ''' +[[All-PyLint-rule-disables-MUST-be-documented-in-the-code]] [cols="2v,v"] |=== -| **Rule** +| <<All-PyLint-rule-disables-MUST-be-documented-in-the-code, Rule>> | All PyLint rule disables MUST be documented in the code. |=== @@ -124,9 +132,10 @@ metadata[line] = results.pop() === Yaml Files (Playbooks, Roles, Vars, etc) ''' +[[Ansible-files-SHOULD-NOT-use-JSON-use-pure-YAML-instead]] [cols="2v,v"] |=== -| **Rule** +| <<Ansible-files-SHOULD-NOT-use-JSON-use-pure-YAML-instead, Rule>> | Ansible files SHOULD NOT use JSON (use pure YAML instead). |=== @@ -144,9 +153,10 @@ Every effort should be made to keep our Ansible YAML files in pure YAML. === Modules ''' +[[Custom-Ansible-modules-SHOULD-be-embedded-in-a-role]] [cols="2v,v"] |=== -| **Rule** +| <<Custom-Ansible-modules-SHOULD-be-embedded-in-a-role, Rule>> | Custom Ansible modules SHOULD be embedded in a role. |=== @@ -177,9 +187,10 @@ The purpose of this rule is to make it easy to include custom modules in our pla ''' +[[Parameters-to-Ansible-modules-SHOULD-use-the-Yaml-dictionary-format-when-3-or-more-parameters-are-being-passed]] [cols="2v,v"] |=== -| **Rule** +| <<Parameters-to-Ansible-modules-SHOULD-use-the-Yaml-dictionary-format-when-3-or-more-parameters-are-being-passed, Rule>> | Parameters to Ansible modules SHOULD use the Yaml dictionary format when 3 or more parameters are being passed |=== @@ -204,9 +215,10 @@ When a module has several parameters that are being passed in, it's hard to see ''' +[[Parameters-to-Ansible-modules-SHOULD-use-the-Yaml-dictionary-format-when-the-line-length-exceeds-120-characters]] [cols="2v,v"] |=== -| **Rule** +| <<Parameters-to-Ansible-modules-SHOULD-use-the-Yaml-dictionary-format-when-the-line-length-exceeds-120-characters, Rule>> | Parameters to Ansible modules SHOULD use the Yaml dictionary format when the line length exceeds 120 characters |=== @@ -228,9 +240,10 @@ Lines that are long quickly become a wall of text that isn't easily parsable. It ---- ''' +[[The-Ansible-command-module-SHOULD-be-used-instead-of-the-Ansible-shell-module]] [cols="2v,v"] |=== -| **Rule** +| <<The-Ansible-command-module-SHOULD-be-used-instead-of-the-Ansible-shell-module, Rule>> | The Ansible `command` module SHOULD be used instead of the Ansible `shell` module. |=== .Context @@ -251,9 +264,10 @@ The Ansible `shell` module can run most commands that can be run from a bash CLI ---- ''' +[[The-Ansible-quote-filter-MUST-be-used-with-any-variable-passed-into-the-shell-module]] [cols="2v,v"] |=== -| **Rule** +| <<The-Ansible-quote-filter-MUST-be-used-with-any-variable-passed-into-the-shell-module, Rule>> | The Ansible `quote` filter MUST be used with any variable passed into the shell module. |=== .Context @@ -279,9 +293,10 @@ It is recommended not to use the `shell` module. However, if it absolutely must * http://docs.ansible.com/fail_module.html[Ansible Fail Module] ''' +[[Ansible-playbooks-MUST-begin-with-checks-for-any-variables-that-they-require]] [cols="2v,v"] |=== -| **Rule** +| <<Ansible-playbooks-MUST-begin-with-checks-for-any-variables-that-they-require, Rule>> | Ansible playbooks MUST begin with checks for any variables that they require. |=== @@ -299,9 +314,10 @@ If an Ansible playbook requires certain variables to be set, it's best to check ---- ''' +[[Ansible-roles-tasks-main-yml-file-MUST-begin-with-checks-for-any-variables-that-they-require]] [cols="2v,v"] |=== -| **Rule** +| <<Ansible-roles-tasks-main-yml-file-MUST-begin-with-checks-for-any-variables-that-they-require, Rule>> | Ansible roles tasks/main.yml file MUST begin with checks for any variables that they require. |=== @@ -318,9 +334,10 @@ If an Ansible role requires certain variables to be set, it's best to check for === Tasks ''' +[[Ansible-tasks-SHOULD-NOT-be-used-in-ansible-playbooks-Instead-use-pre_tasks-and-post_tasks]] [cols="2v,v"] |=== -| **Rule** +| <<Ansible-tasks-SHOULD-NOT-be-used-in-ansible-playbooks-Instead-use-pre_tasks-and-post_tasks, Rule>> | Ansible tasks SHOULD NOT be used in ansible playbooks. Instead, use pre_tasks and post_tasks. |=== An Ansible play is defined as a Yaml dictionary. Because of that, ansible doesn't know if the play's tasks list or roles list was specified first. Therefore Ansible always runs tasks after roles. @@ -370,9 +387,10 @@ Therefore, we SHOULD use pre_tasks and post_tasks to make it more clear when the === Roles ''' +[[All-tasks-in-a-role-SHOULD-be-tagged-with-the-role-name]] [cols="2v,v"] |=== -| **Rule** +| <<All-tasks-in-a-role-SHOULD-be-tagged-with-the-role-name, Rule>> | All tasks in a role SHOULD be tagged with the role name. |=== @@ -395,9 +413,10 @@ This is very useful when developing and debugging new tasks. It can also signifi ''' +[[The-Ansible-roles-directory-MUST-maintain-a-flat-structure]] [cols="2v,v"] |=== -| **Rule** +| <<The-Ansible-roles-directory-MUST-maintain-a-flat-structure, Rule>> | The Ansible roles directory MUST maintain a flat structure. |=== @@ -410,9 +429,10 @@ This is very useful when developing and debugging new tasks. It can also signifi * Make it compatible with Ansible Galaxy ''' +[[Ansible-Roles-SHOULD-be-named-like-technology_component_subcomponent]] [cols="2v,v"] |=== -| **Rule** +| [[Ansible-Roles-SHOULD-be-named-like-technology_component_subcomponent, Rule]] | Ansible Roles SHOULD be named like technology_component[_subcomponent]. |=== @@ -430,9 +450,10 @@ Many times the `technology` portion of the pattern will line up with a package n * http://jinja.pocoo.org/docs/dev/templates/#builtin-filters[Jinja2 Builtin Filters] ''' +[[The-default-filter-SHOULD-replace-empty-strings-lists-etc]] [cols="2v,v"] |=== -| **Rule** +| <<The-default-filter-SHOULD-replace-empty-strings-lists-etc, Rule>> | The `default` filter SHOULD replace empty strings, lists, etc. |=== @@ -469,15 +490,17 @@ This is almost always more desirable than an empty list, string, etc. === Yum and DNF ''' +[[Package-installation-MUST-use-ansible-action-module-to-abstract-away-dnf-yum]] [cols="2v,v"] |=== -| **Rule** +| <<Package-installation-MUST-use-ansible-action-module-to-abstract-away-dnf-yum, Rule>> | Package installation MUST use ansible action module to abstract away dnf/yum. -| Package installation MUST use name= and state=present rather than pkg= and state=installed respectively. |=== + +[[Package-installation-MUST-use-name-and-state-present-rather-than-pkg-and-state-installed-respectively]] [cols="2v,v"] |=== -| **Rule** +| <<Package-installation-MUST-use-name-and-state-present-rather-than-pkg-and-state-installed-respectively, Rule>> | Package installation MUST use name= and state=present rather than pkg= and state=installed respectively. |=== diff --git a/docs/style_guide.adoc b/docs/style_guide.adoc index 09d4839c7..72eaedcf9 100644 --- a/docs/style_guide.adoc +++ b/docs/style_guide.adoc @@ -19,9 +19,10 @@ This style guide complies with https://www.ietf.org/rfc/rfc2119.txt[RFC2119]. * https://www.python.org/dev/peps/pep-0008/#maximum-line-length[Python Pep8 Line Length] ''' +[[All-lines-SHOULD-be-no-longer-than-80-characters]] [cols="2v,v"] |=== -| **Rule** +| <<All-lines-SHOULD-be-no-longer-than-80-characters, Rule>> | All lines SHOULD be no longer than 80 characters. |=== @@ -31,9 +32,10 @@ Code readability is subjective, therefore pull-requests SHOULD still be merged, ''' +[[All-lines-MUST-be-no-longer-than-120-characters]] [cols="2v,v"] |=== -| **Rule** +| <<All-lines-MUST-be-no-longer-than-120-characters, Rule>> | All lines MUST be no longer than 120 characters. |=== @@ -46,9 +48,10 @@ This is a hard limit and is enforced by the build bot. This check MUST NOT be di === Ansible Yaml file extension ''' +[[All-Ansible-Yaml-files-MUST-have-a-yml-extension-and-NOT-YML-yaml-etc]] [cols="2v,v"] |=== -| **Rule** +| <<All-Ansible-Yaml-files-MUST-have-a-yml-extension-and-NOT-YML-yaml-etc, Rule>> | All Ansible Yaml files MUST have a .yml extension (and NOT .YML, .yaml etc). |=== @@ -59,9 +62,10 @@ Example: `tasks.yml` === Ansible CLI Variables ''' +[[Variables-meant-to-be-passed-in-from-the-ansible-CLI-MUST-have-a-prefix-of-cli]] [cols="2v,v"] |=== -| **Rule** +| <<Variables-meant-to-be-passed-in-from-the-ansible-CLI-MUST-have-a-prefix-of-cli, Rule>> | Variables meant to be passed in from the ansible CLI MUST have a prefix of cli_ |=== @@ -76,9 +80,10 @@ ansible-playbook -e cli_foo=bar someplays.yml === Ansible Global Variables ''' +[[Global-variables-MUST-have-a-prefix-of-g]] [cols="2v,v"] |=== -| **Rule** +| <<Global-variables-MUST-have-a-prefix-of-g, Rule>> | Global variables MUST have a prefix of g_ |=== Ansible global variables are defined as any variables outside of ansible roles. Examples include playbook variables, variables passed in on the cli, etc. @@ -94,9 +99,10 @@ g_environment: someval Ansible role variables are defined as variables contained in (or passed into) a role. ''' +[[Role-variables-MUST-have-a-prefix-of-atleast-3-characters-See.below.for.specific.naming.rules]] [cols="2v,v"] |=== -| **Rule** +| <<Role-variables-MUST-have-a-prefix-of-atleast-3-characters-See.below.for.specific.naming.rules, Rule>> | Role variables MUST have a prefix of atleast 3 characters. See below for specific naming rules. |=== diff --git a/filter_plugins/openshift_master.py b/filter_plugins/openshift_master.py index 8d7c62ad1..7ababc51a 100644 --- a/filter_plugins/openshift_master.py +++ b/filter_plugins/openshift_master.py @@ -463,6 +463,32 @@ class FilterModule(object): IdentityProviderBase.validate_idp_list(idp_list) return yaml.safe_dump([idp.to_dict() for idp in idp_list], default_flow_style=False) + @staticmethod + def validate_pcs_cluster(data, masters=None): + ''' Validates output from "pcs status", ensuring that each master + provided is online. + Ex: data = ('...', + 'PCSD Status:', + 'master1.example.com: Online', + 'master2.example.com: Online', + 'master3.example.com: Online', + '...') + masters = ['master1.example.com', + 'master2.example.com', + 'master3.example.com'] + returns True + ''' + if not issubclass(type(data), str): + raise errors.AnsibleFilterError("|failed expects data is a string") + if not issubclass(type(masters), list): + raise errors.AnsibleFilterError("|failed expects masters is a list") + valid = True + for master in masters: + if "{0}: Online".format(master) not in data: + valid = False + return valid + def filters(self): ''' returns a mapping of filters to methods ''' - return {"translate_idps": self.translate_idps} + return {"translate_idps": self.translate_idps, + "validate_pcs_cluster": self.validate_pcs_cluster} diff --git a/git/parent.py b/git/parent.py new file mode 100755 index 000000000..154a02350 --- /dev/null +++ b/git/parent.py @@ -0,0 +1,96 @@ +#!/usr/bin/env python +''' + Script to determine if this commit has also + been merged through the stage branch +''' +# +# Usage: +# parent_check.py <branch> <commit_id> +# +# +import sys +import subprocess + +def run_cli_cmd(cmd, in_stdout=None, in_stderr=None): + '''Run a command and return its output''' + if not in_stderr: + proc = subprocess.Popen(cmd, bufsize=-1, stderr=subprocess.PIPE, stdout=subprocess.PIPE, shell=False) + else: + proc = subprocess.check_output(cmd, bufsize=-1, stdout=in_stdout, stderr=in_stderr, shell=False) + stdout, stderr = proc.communicate() + if proc.returncode != 0: + return {"rc": proc.returncode, "error": stderr} + else: + return {"rc": proc.returncode, "result": stdout} + +def main(): + '''Check to ensure that the commit that is currently + being submitted is also in the stage branch. + + if it is, succeed + else, fail + ''' + branch = 'prod' + + if sys.argv[1] != branch: + sys.exit(0) + + # git co stg + results = run_cli_cmd(['/usr/bin/git', 'checkout', 'stg']) + + # git pull latest + results = run_cli_cmd(['/usr/bin/git', 'pull']) + + # setup on the <prod> branch in git + results = run_cli_cmd(['/usr/bin/git', 'checkout', 'prod']) + + results = run_cli_cmd(['/usr/bin/git', 'pull']) + # merge the passed in commit into my current <branch> + + commit_id = sys.argv[2] + results = run_cli_cmd(['/usr/bin/git', 'merge', commit_id]) + + # get the differences from stg and <branch> + results = run_cli_cmd(['/usr/bin/git', 'rev-list', '--left-right', 'stg...prod']) + + # exit here with error code if the result coming back is an error + if results['rc'] != 0: + print results['error'] + sys.exit(results['rc']) + + count = 0 + # Each 'result' is a commit + # Walk through each commit and see if it is in stg + for commit in results['result'].split('\n'): + + # continue if it is already in stg + if not commit or commit.startswith('<'): + continue + + # remove the first char '>' + commit = commit[1:] + + # check if any remote branches contain $commit + results = run_cli_cmd(['/usr/bin/git', 'branch', '-q', '-r', '--contains', commit], in_stderr=None) + + # if this comes back empty, nothing contains it, we can skip it as + # we have probably created the merge commit here locally + if results['rc'] == 0 and len(results['result']) == 0: + continue + + # The results generally contain origin/pr/246/merge and origin/pr/246/head + # this is the pull request which would contain the commit in question. + # + # If the results do not contain origin/stg then stage does not contain + # the commit in question. Therefore we need to alert! + if 'origin/stg' not in results['result']: + print "\nFAILED: (These commits are not in stage.)\n" + print "\t%s" % commit + count += 1 + + # Exit with count of commits in #{branch} but not stg + sys.exit(count) + +if __name__ == '__main__': + main() + diff --git a/git/parent.rb b/git/parent.rb deleted file mode 100755 index 2acb127c4..000000000 --- a/git/parent.rb +++ /dev/null @@ -1,45 +0,0 @@ -#!/usr/bin/env ruby -# -# -# - -if __FILE__ == $0 - # If we aren't on master we don't need to parent check - branch = 'prod' - exit(0) if ARGV[0] !~ /#{branch}/ - commit_id = ARGV[1] - %x[/usr/bin/git checkout #{branch}] - %x[/usr/bin/git merge #{commit_id}] - - count = 0 - #lines = %x[/usr/bin/git rev-list --left-right stg...master].split("\n") - lines = %x[/usr/bin/git rev-list --left-right remotes/origin/stg...#{branch}].split("\n") - lines.each do |commit| - # next if they are in stage - next if commit =~ /^</ - # remove the first char '>' - commit = commit[1..-1] - # check if any remote branches contain $commit - results = %x[/usr/bin/git branch -q -r --contains #{commit} 2>/dev/null ] - # if this comes back empty, nothing contains it, we can skip it as - # we have probably created the merge commit here locally - next if results.empty? - - # The results generally contain origin/pr/246/merge and origin/pr/246/head - # this is the pull request which would contain the commit in question. - # - # If the results do not contain origin/stg then stage does not contain - # the commit in question. Therefore we need to alert! - unless results =~ /origin\/stg/ - puts "\nFAILED: (These commits are not in stage.)\n" - puts "\t#{commit}" - count += 1 - end - end - - # Exit with count of commits in #{branch} but not stg - exit(count) -end - -__END__ - diff --git a/git/yaml_validation.py b/git/yaml_validation.py new file mode 100755 index 000000000..94b8b0435 --- /dev/null +++ b/git/yaml_validation.py @@ -0,0 +1,68 @@ +#!/usr/bin/env python +# +# python yaml validator for a git commit +# +''' +python yaml validator for a git commit +''' +import shutil +import sys +import os +import tempfile +import subprocess +import yaml + +def get_changes(oldrev, newrev, tempdir): + '''Get a list of git changes from oldrev to newrev''' + proc = subprocess.Popen(['/usr/bin/git', 'diff', '--name-only', oldrev, + newrev, '--diff-filter=ACM'], stdout=subprocess.PIPE) + stdout, _ = proc.communicate() + files = stdout.split('\n') + + # No file changes + if not files: + return [] + + cmd = '/usr/bin/git archive %s %s | /bin/tar x -C %s' % (newrev, " ".join(files), tempdir) + proc = subprocess.Popen(cmd, shell=True) + _, _ = proc.communicate() + + rfiles = [] + for dirpath, _, fnames in os.walk(tempdir): + for fname in fnames: + rfiles.append(os.path.join(dirpath, fname)) + + return rfiles + +def main(): + ''' + Perform yaml validation + ''' + results = [] + try: + tmpdir = tempfile.mkdtemp(prefix='jenkins-git-') + old, new, _ = sys.argv[1:] + + for file_mod in get_changes(old, new, tmpdir): + + print "+++++++ Received: %s" % file_mod + + if not file_mod.endswith('.yml') and not file_mod.endswith('.yaml'): + continue + + try: + yaml.load(open(file_mod)) + results.append(True) + + except yaml.scanner.ScannerError as yerr: + print yerr + results.append(False) + finally: + shutil.rmtree(tmpdir) + + if not all(results): + sys.exit(1) + +if __name__ == "__main__": + main() + diff --git a/git/yaml_validation.rb b/git/yaml_validation.rb deleted file mode 100755 index f5ded7a78..000000000 --- a/git/yaml_validation.rb +++ /dev/null @@ -1,72 +0,0 @@ -#!/usr/bin/env ruby -# -# -# -require 'yaml' -require 'tmpdir' - -class YamlValidate - def self.yaml_file?(filename) - return filename.end_with?('.yaml') || filename.end_with?('.yml') - end - - def self.short_yaml_ext?(filename) - return filename.end_with?(".yml") - end - - def self.valid_yaml?(filename) - YAML::load_file(filename) - - return true - end -end - -class GitCommit - attr_accessor :oldrev, :newrev, :refname, :tmp - def initialize(oldrev, newrev, refname) - @oldrev = oldrev - @newrev = newrev - @refname = refname - @tmp = Dir.mktmpdir(@newrev) - end - - def get_file_changes() - files = %x[/usr/bin/git diff --name-only #{@oldrev} #{@newrev} --diff-filter=ACM].split("\n") - - # if files is empty we will get a full checkout. This happens on - # a git rm file. If there are no changes then we need to skip the archive - return [] if files.empty? - - # We only want to take the files that changed. Archive will do that when passed - # the filenames. It will export these to a tmp dir - system("/usr/bin/git archive #{@newrev} #{files.join(" ")} | tar x -C #{@tmp}") - return Dir.glob("#{@tmp}/**/*").delete_if { |file| File.directory?(file) } - end -end - -if __FILE__ == $0 - while data = STDIN.gets - oldrev, newrev, refname = data.split - gc = GitCommit.new(oldrev, newrev, refname) - - results = [] - gc.get_file_changes().each do |file| - begin - puts "++++++ Received: #{file}" - - #raise "Yaml file extensions must be .yaml not .yml" if YamlValidate.short_yaml_ext? file - - # skip readme, other files, etc - next unless YamlValidate.yaml_file?(file) - - results << YamlValidate.valid_yaml?(file) - rescue Exception => ex - puts "\n#{ex.message}\n\n" - results << false - end - end - - #puts "RESULTS\n#{results.inspect}\n" - exit 1 if results.include?(false) - end -end diff --git a/inventory/aws/hosts/ec2.ini b/inventory/aws/hosts/ec2.ini index 1f503b8cf..aa0f9090f 100644 --- a/inventory/aws/hosts/ec2.ini +++ b/inventory/aws/hosts/ec2.ini @@ -45,10 +45,10 @@ vpc_destination_variable = ip_address route53 = False # To exclude RDS instances from the inventory, uncomment and set to False. -#rds = False +rds = False # To exclude ElastiCache instances from the inventory, uncomment and set to False. -#elasticache = False +elasticache = False # Additionally, you can specify the list of zones to exclude looking up in # 'route53_excluded_zones' as a comma-separated list. diff --git a/openshift-ansible.spec b/openshift-ansible.spec index 70938e8d2..4d00c655b 100644 --- a/openshift-ansible.spec +++ b/openshift-ansible.spec @@ -5,7 +5,7 @@ } Name: openshift-ansible -Version: 3.0.28 +Version: 3.0.35 Release: 1%{?dist} Summary: Openshift and Atomic Enterprise Ansible License: ASL 2.0 @@ -259,6 +259,77 @@ Atomic OpenShift Utilities includes %changelog +* Mon Jan 18 2016 Brenton Leanhardt <bleanhar@redhat.com> 3.0.35-1 +- added the lib_timedate role (mwoodson@redhat.com) +- added chrony (mwoodson@redhat.com) +- added oso_moniotoring tools role (mwoodson@redhat.com) +- Improve pacemaker 'is-active' check. (abutcher@redhat.com) + +* Mon Jan 18 2016 Brenton Leanhardt <bleanhar@redhat.com> 3.0.34-1 +- clean up too-many-branches / logic (jdiaz@redhat.com) +- atomic-openshift-installer: add containerized to inventory + (smunilla@redhat.com) +- Add 'unknown' to possible output for the is-active check. + (abutcher@redhat.com) +- Fix cluster_method conditional in master restart playbook. + (abutcher@redhat.com) +- Use IdentityFile instead of PrivateKey (donovan.muller@gmail.com) +- atomic-openshift-installer: Remove containerized install for 3.0 + (smunilla@redhat.com) +- Host group should be OSEv3 not OSv3 (donovan.muller@gmail.com) +- Remove pause after haproxy start (abutcher@redhat.com) +- Ensure nfs-utils installed for non-atomic hosts. (abutcher@redhat.com) + +* Fri Jan 15 2016 Brenton Leanhardt <bleanhar@redhat.com> 3.0.33-1 +- Configure nodes which are also masters prior to nodes in containerized + install. (abutcher@redhat.com) +- Call attention to openshift_master_rolling_restart_mode variable in restart + prompt. (abutcher@redhat.com) +- Added anchors for rules in style_guide.adoc in order to make it easier to + reference specific rules in PRs. (twiest@redhat.com) +- Update ec2.ini (jdetiber@redhat.com) + +* Thu Jan 14 2016 Brenton Leanhardt <bleanhar@redhat.com> 3.0.32-1 +- Uninstall remove containerized wrapper and symlinks (abutcher@redhat.com) + +* Thu Jan 14 2016 Brenton Leanhardt <bleanhar@redhat.com> 3.0.31-1 +- Check api prior to starting node. (abutcher@redhat.com) +- added anchors (twiest@redhat.com) + +* Wed Jan 13 2016 Joel Diaz <jdiaz@redhat.com> 3.0.30-1 +- Add -A and detail --v3 flags + +* Wed Jan 13 2016 Brenton Leanhardt <bleanhar@redhat.com> 3.0.29-1 +- 3.1.1 upgrade playbook (bleanhar@redhat.com) +- Updated help menu for v3 flag (kwoodson@redhat.com) +- Add wait in between api and controllers start for native ha. + (abutcher@redhat.com) +- atomic-openshift-installer: Error handling for unicode hostnames + (smunilla@redhat.com) +- Update api verification. (abutcher@redhat.com) +- Add a Verify API Server handler that waits for the API server to become + available (sdodson@redhat.com) +- Add -A parameter to forward ssh agent (jdiaz@redhat.com) +- Validate pacemaker cluster members. (abutcher@redhat.com) +- Removed atomic host check (kwoodson@redhat.com) +- Add is_containerized inputs to nosetests. (abutcher@redhat.com) +- Add wait for API before starting controllers w/ native ha install. + (abutcher@redhat.com) +- Fix for to_padded_yaml filter (jdetiber@redhat.com) +- - sqashed to one commit (llange@redhat.com) +- Switch to using hostnamectl as it works on atomic and rhel7 + (sdodson@redhat.com) +- Update rolling restart playbook for pacemaker support. Replace fail with a + warn and prompt if running ansible from a host that will be rebooted. Re- + organize playbooks. (abutcher@redhat.com) +- Implement simple master rolling restarts. (dgoodwin@redhat.com) +- re-enable containerize installs (sdodson@redhat.com) +- Set portal net in master playbook (jdetiber@redhat.com) +- Set the cli image to match osm_image in openshift_cli role + (sdodson@redhat.com) +- atomic-openshift-installer: Populate new_nodes group (smunilla@redhat.com) +- Always pull docker images (sdodson@redhat.com) + * Mon Jan 11 2016 Kenny Woodson <kwoodson@redhat.com> 3.0.28-1 - added the rhe7-host-monitoring service file (mwoodson@redhat.com) - Fixing tab completion for latest metadata changes (kwoodson@redhat.com) diff --git a/playbooks/adhoc/bootstrap-fedora.yml b/playbooks/adhoc/bootstrap-fedora.yml index de9f36c8a..0df77e309 100644 --- a/playbooks/adhoc/bootstrap-fedora.yml +++ b/playbooks/adhoc/bootstrap-fedora.yml @@ -1,4 +1,4 @@ -- hosts: OSv3 +- hosts: OSEv3 gather_facts: false tasks: - name: install python and deps for ansible modules diff --git a/playbooks/adhoc/grow_docker_vg/grow_docker_vg.yml b/playbooks/adhoc/grow_docker_vg/grow_docker_vg.yml index 174cea460..d24e9cafa 100644 --- a/playbooks/adhoc/grow_docker_vg/grow_docker_vg.yml +++ b/playbooks/adhoc/grow_docker_vg/grow_docker_vg.yml @@ -20,7 +20,7 @@ # ansible-playbook -e 'cli_tag_name=ops-compute-12345' grow_docker_vg.yml # # Notes: -# * By default this will do a 55GB GP2 volume. The can be overidden with the "-e 'cli_volume_size=100'" variable +# * By default this will do a 200GB GP2 volume. The can be overidden with the "-e 'cli_volume_size=100'" variable # * This does a GP2 by default. Support for Provisioned IOPS has not been added # * This will assign the new volume to /dev/xvdc. This is not variablized, yet. # * This can be done with NO downtime on the host @@ -36,7 +36,7 @@ vars: cli_volume_type: gp2 - cli_volume_size: 55 + cli_volume_size: 200 # cli_volume_iops: "{{ 30 * cli_volume_size }}" pre_tasks: diff --git a/playbooks/adhoc/uninstall.yml b/playbooks/adhoc/uninstall.yml index ac20f5f9b..36d686c8b 100644 --- a/playbooks/adhoc/uninstall.yml +++ b/playbooks/adhoc/uninstall.yml @@ -202,6 +202,10 @@ - /usr/lib/systemd/system/atomic-openshift-master-controllers.service - /usr/lib/systemd/system/origin-master-api.service - /usr/lib/systemd/system/origin-master-controllers.service + - /usr/local/bin/openshift + - /usr/local/bin/oadm + - /usr/local/bin/oc + - /usr/local/bin/kubectl # Since we are potentially removing the systemd unit files for separated # master-api and master-controllers services, so we need to reload the diff --git a/playbooks/byo/openshift-cluster/upgrades/v3_1_minor/README.md b/playbooks/byo/openshift-cluster/upgrades/v3_1_minor/README.md new file mode 100644 index 000000000..b230835c3 --- /dev/null +++ b/playbooks/byo/openshift-cluster/upgrades/v3_1_minor/README.md @@ -0,0 +1,17 @@ +# v3.1 minor upgrade playbook +This upgrade will preserve all locally made configuration modifications to the +Masters and Nodes. + +## Overview +This playbook is available as a technical preview. It currently performs the +following steps. + + * Upgrade and restart master services + * Upgrade and restart node services + * Applies the latest cluster policies + * Updates the default router if one exists + * Updates the default registry if one exists + * Updates image streams and quickstarts + +## Usage +ansible-playbook -i ~/ansible-inventory openshift-ansible/playbooks/byo/openshift-cluster/upgrades/v3_1_minor/upgrade.yml diff --git a/playbooks/byo/openshift-cluster/upgrades/v3_1_minor/upgrade.yml b/playbooks/byo/openshift-cluster/upgrades/v3_1_minor/upgrade.yml new file mode 100644 index 000000000..20fa9b10f --- /dev/null +++ b/playbooks/byo/openshift-cluster/upgrades/v3_1_minor/upgrade.yml @@ -0,0 +1,14 @@ +--- +- include: ../../../../common/openshift-cluster/evaluate_groups.yml + vars: + g_etcd_hosts: "{{ groups.etcd | default([]) }}" + g_master_hosts: "{{ groups.masters | default([]) }}" + g_nfs_hosts: "{{ groups.nfs | default([]) }}" + g_node_hosts: "{{ groups.nodes | default([]) }}" + g_lb_hosts: "{{ groups.lb | default([]) }}" + openshift_cluster_id: "{{ cluster_id | default('default') }}" + openshift_deployment_type: "{{ deployment_type }}" +- include: ../../../../common/openshift-cluster/upgrades/v3_1_minor/pre.yml +- include: ../../../../common/openshift-cluster/upgrades/v3_1_minor/upgrade.yml +- include: ../../../openshift-master/restart.yml +- include: ../../../../common/openshift-cluster/upgrades/v3_1_minor/post.yml diff --git a/playbooks/byo/openshift-master/filter_plugins b/playbooks/byo/openshift-master/filter_plugins new file mode 120000 index 000000000..99a95e4ca --- /dev/null +++ b/playbooks/byo/openshift-master/filter_plugins @@ -0,0 +1 @@ +../../../filter_plugins
\ No newline at end of file diff --git a/playbooks/byo/openshift-master/lookup_plugins b/playbooks/byo/openshift-master/lookup_plugins new file mode 120000 index 000000000..ac79701db --- /dev/null +++ b/playbooks/byo/openshift-master/lookup_plugins @@ -0,0 +1 @@ +../../../lookup_plugins
\ No newline at end of file diff --git a/playbooks/byo/openshift-master/restart.yml b/playbooks/byo/openshift-master/restart.yml new file mode 100644 index 000000000..a78a6aa3d --- /dev/null +++ b/playbooks/byo/openshift-master/restart.yml @@ -0,0 +1,4 @@ +--- +- include: ../../common/openshift-master/restart.yml + vars_files: + - ../../byo/openshift-cluster/cluster_hosts.yml diff --git a/playbooks/byo/openshift-master/roles b/playbooks/byo/openshift-master/roles new file mode 120000 index 000000000..20c4c58cf --- /dev/null +++ b/playbooks/byo/openshift-master/roles @@ -0,0 +1 @@ +../../../roles
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/v3_1_minor/filter_plugins b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/filter_plugins new file mode 120000 index 000000000..27ddaa18b --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/filter_plugins @@ -0,0 +1 @@ +../../../../../filter_plugins
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/v3_1_minor/library b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/library new file mode 120000 index 000000000..53bed9684 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/library @@ -0,0 +1 @@ +../library
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/v3_1_minor/lookup_plugins b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/lookup_plugins new file mode 120000 index 000000000..cf407f69b --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/lookup_plugins @@ -0,0 +1 @@ +../../../../../lookup_plugins
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/v3_1_minor/post.yml b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/post.yml new file mode 100644 index 000000000..d8336fcae --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/post.yml @@ -0,0 +1,50 @@ +--- +############################################################################### +# Post upgrade - Upgrade default router, default registry and examples +############################################################################### +- name: Upgrade default router and default registry + hosts: oo_first_master + vars: + openshift_deployment_type: "{{ deployment_type }}" + registry_image: "{{ openshift.master.registry_url | replace( '${component}', 'docker-registry' ) | replace ( '${version}', 'v' + g_new_version ) }}" + router_image: "{{ openshift.master.registry_url | replace( '${component}', 'haproxy-router' ) | replace ( '${version}', 'v' + g_new_version ) }}" + oc_cmd: "{{ openshift.common.client_binary }} --config={{ openshift.common.config_base }}/master/admin.kubeconfig" + roles: + # Create the new templates shipped in 3.1.z, existing templates are left + # unmodified. This prevents the subsequent role definition for + # openshift_examples from failing when trying to replace templates that do + # not already exist. We could have potentially done a replace --force to + # create and update in one step. + - openshift_examples + # Update the existing templates + - role: openshift_examples + openshift_examples_import_command: replace + pre_tasks: + - name: Check for default router + command: > + {{ oc_cmd }} get -n default dc/router + register: _default_router + failed_when: false + changed_when: false + + - name: Check for default registry + command: > + {{ oc_cmd }} get -n default dc/docker-registry + register: _default_registry + failed_when: false + changed_when: false + + - name: Update router image to current version + when: _default_router.rc == 0 + command: > + {{ oc_cmd }} patch dc/router -p + '{"spec":{"template":{"spec":{"containers":[{"name":"router","image":"{{ router_image }}"}]}}}}' + --api-version=v1 + + - name: Update registry image to current version + when: _default_registry.rc == 0 + command: > + {{ oc_cmd }} patch dc/docker-registry -p + '{"spec":{"template":{"spec":{"containers":[{"name":"registry","image":"{{ registry_image }}"}]}}}}' + --api-version=v1 + diff --git a/playbooks/common/openshift-cluster/upgrades/v3_1_minor/pre.yml b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/pre.yml new file mode 100644 index 000000000..91780de09 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/pre.yml @@ -0,0 +1,87 @@ +--- +############################################################################### +# Evaluate host groups and gather facts +############################################################################### +- name: Load openshift_facts + hosts: oo_masters_to_config:oo_nodes_to_config:oo_etcd_to_config:oo_lb_to_config + roles: + - openshift_facts + +############################################################################### +# Pre-upgrade checks +############################################################################### +- name: Verify upgrade can proceed + hosts: oo_first_master + vars: + openshift_master_ha: "{{ groups.oo_masters_to_config | length > 1 }}" + target_version: "{{ '1.1.1' if deployment_type == 'origin' else '3.1.1' }}" + gather_facts: no + tasks: + - fail: + msg: > + This upgrade is only supported for origin, openshift-enterprise, and online + deployment types + when: deployment_type not in ['origin','openshift-enterprise', 'online'] + + - fail: + msg: > + openshift_pkg_version is {{ openshift_pkg_version }} which is not a + valid version for a {{ target_version }} upgrade + when: openshift_pkg_version is defined and openshift_pkg_version.split('-',1).1 | version_compare(target_version ,'<') + +- name: Verify upgrade can proceed + hosts: oo_masters_to_config:oo_nodes_to_config + vars: + target_version: "{{ '1.1.1' if deployment_type == 'origin' else '3.1.1' }}" + tasks: + - name: Clean package cache + command: "{{ ansible_pkg_mgr }} clean all" + + - set_fact: + g_new_service_name: "{{ 'origin' if deployment_type =='origin' else 'atomic-openshift' }}" + + - name: Determine available versions + script: ../files/versions.sh {{ g_new_service_name }} openshift + register: g_versions_result + + - set_fact: + g_aos_versions: "{{ g_versions_result.stdout | from_yaml }}" + + - set_fact: + g_new_version: "{{ g_aos_versions.curr_version.split('-', 1).0 if g_aos_versions.avail_version is none else g_aos_versions.avail_version.split('-', 1).0 }}" + + - fail: + msg: This playbook requires Origin 1.1 or later + when: deployment_type == 'origin' and g_aos_versions.curr_version | version_compare('1.1','<') + + - fail: + msg: This playbook requires Atomic Enterprise Platform/OpenShift Enterprise 3.1 or later + when: deployment_type == 'atomic-openshift' and g_aos_versions.curr_version | version_compare('3.1','<') + + - fail: + msg: Upgrade packages not found + when: (g_aos_versions.avail_version | default(g_aos_versions.curr_version, true) | version_compare(target_version, '<')) + + - set_fact: + pre_upgrade_complete: True + + +############################################################################## +# Gate on pre-upgrade checks +############################################################################## +- name: Gate on pre-upgrade checks + hosts: localhost + connection: local + become: no + vars: + pre_upgrade_hosts: "{{ groups.oo_masters_to_config | union(groups.oo_nodes_to_config) }}" + tasks: + - set_fact: + pre_upgrade_completed: "{{ hostvars + | oo_select_keys(pre_upgrade_hosts) + | oo_collect('inventory_hostname', {'pre_upgrade_complete': true}) }}" + - set_fact: + pre_upgrade_failed: "{{ pre_upgrade_hosts | difference(pre_upgrade_completed) }}" + - fail: + msg: "Upgrade cannot continue. The following hosts did not complete pre-upgrade checks: {{ pre_upgrade_failed | join(',') }}" + when: pre_upgrade_failed | length > 0 diff --git a/playbooks/common/openshift-cluster/upgrades/v3_1_minor/roles b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/roles new file mode 120000 index 000000000..6bc1a7aef --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/roles @@ -0,0 +1 @@ +../../../../../roles
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/v3_1_minor/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/upgrade.yml new file mode 100644 index 000000000..81dbba1e3 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_1_minor/upgrade.yml @@ -0,0 +1,137 @@ +--- +############################################################################### +# The restart playbook should be run after this playbook completes. +############################################################################### + +############################################################################### +# Upgrade Masters +############################################################################### +- name: Upgrade master packages and configuration + hosts: oo_masters_to_config + vars: + openshift_version: "{{ openshift_pkg_version | default('') }}" + tasks: + - name: Upgrade master packages + command: "{{ ansible_pkg_mgr}} update -y {{ openshift.common.service_type }}-master{{ openshift_version }}" + + - name: Ensure python-yaml present for config upgrade + action: "{{ ansible_pkg_mgr }} name=PyYAML state=present" + when: not openshift.common.is_atomic | bool + +# Currently 3.1.1 does not have any new configuration settings +# +# - name: Upgrade master configuration +# openshift_upgrade_config: +# from_version: '3.0' +# to_version: '3.1' +# role: master +# config_base: "{{ hostvars[inventory_hostname].openshift.common.config_base }}" + +- name: Set master update status to complete + hosts: oo_masters_to_config + tasks: + - set_fact: + master_update_complete: True + +############################################################################## +# Gate on master update complete +############################################################################## +- name: Gate on master update + hosts: localhost + connection: local + become: no + tasks: + - set_fact: + master_update_completed: "{{ hostvars + | oo_select_keys(groups.oo_masters_to_config) + | oo_collect('inventory_hostname', {'master_update_complete': true}) }}" + - set_fact: + master_update_failed: "{{ groups.oo_masters_to_config | difference(master_update_completed) }}" + - fail: + msg: "Upgrade cannot continue. The following masters did not finish updating: {{ master_update_failed | join(',') }}" + when: master_update_failed | length > 0 + +############################################################################### +# Upgrade Nodes +############################################################################### +- name: Upgrade nodes + hosts: oo_nodes_to_config + vars: + openshift_version: "{{ openshift_pkg_version | default('') }}" + roles: + - openshift_facts + tasks: + - name: Upgrade node packages + command: "{{ ansible_pkg_mgr }} update -y {{ openshift.common.service_type }}-node{{ openshift_version }}" + + - name: Restart node service + service: name="{{ openshift.common.service_type }}-node" state=restarted + + - set_fact: + node_update_complete: True + +############################################################################## +# Gate on nodes update +############################################################################## +- name: Gate on nodes update + hosts: localhost + connection: local + become: no + tasks: + - set_fact: + node_update_completed: "{{ hostvars + | oo_select_keys(groups.oo_nodes_to_config) + | oo_collect('inventory_hostname', {'node_update_complete': true}) }}" + - set_fact: + node_update_failed: "{{ groups.oo_nodes_to_config | difference(node_update_completed) }}" + - fail: + msg: "Upgrade cannot continue. The following nodes did not finish updating: {{ node_update_failed | join(',') }}" + when: node_update_failed | length > 0 + +############################################################################### +# Reconcile Cluster Roles and Cluster Role Bindings +############################################################################### +- name: Reconcile Cluster Roles and Cluster Role Bindings + hosts: oo_masters_to_config + vars: + origin_reconcile_bindings: "{{ deployment_type == 'origin' and g_new_version | version_compare('1.0.6', '>') }}" + ent_reconcile_bindings: true + openshift_master_ha: "{{ groups.oo_masters_to_config | length > 1 }}" + tasks: + - name: Reconcile Cluster Roles + command: > + {{ openshift.common.admin_binary}} --config={{ openshift.common.config_base }}/master/admin.kubeconfig + policy reconcile-cluster-roles --confirm + run_once: true + + - name: Reconcile Cluster Role Bindings + command: > + {{ openshift.common.admin_binary}} --config={{ openshift.common.config_base }}/master/admin.kubeconfig + policy reconcile-cluster-role-bindings + --exclude-groups=system:authenticated + --exclude-groups=system:unauthenticated + --exclude-users=system:anonymous + --additive-only=true --confirm + when: origin_reconcile_bindings | bool or ent_reconcile_bindings | bool + run_once: true + + - set_fact: + reconcile_complete: True + +############################################################################## +# Gate on reconcile +############################################################################## +- name: Gate on reconcile + hosts: localhost + connection: local + become: no + tasks: + - set_fact: + reconcile_completed: "{{ hostvars + | oo_select_keys(groups.oo_masters_to_config) + | oo_collect('inventory_hostname', {'reconcile_complete': true}) }}" + - set_fact: + reconcile_failed: "{{ groups.oo_masters_to_config | difference(reconcile_completed) }}" + - fail: + msg: "Upgrade cannot continue. The following masters did not finish reconciling: {{ reconcile_failed | join(',') }}" + when: reconcile_failed | length > 0 diff --git a/playbooks/common/openshift-master/config.yml b/playbooks/common/openshift-master/config.yml index 543583642..296176e49 100644 --- a/playbooks/common/openshift-master/config.yml +++ b/playbooks/common/openshift-master/config.yml @@ -314,6 +314,7 @@ - name: Configure master instances hosts: oo_masters_to_config + any_errors_fatal: true serial: 1 vars: sync_tmpdir: "{{ hostvars.localhost.g_master_mktemp.stdout }}" diff --git a/playbooks/common/openshift-master/restart.yml b/playbooks/common/openshift-master/restart.yml new file mode 100644 index 000000000..052892863 --- /dev/null +++ b/playbooks/common/openshift-master/restart.yml @@ -0,0 +1,151 @@ +--- +- include: ../openshift-cluster/evaluate_groups.yml + +- name: Validate configuration for rolling restart + hosts: oo_masters_to_config + roles: + - openshift_facts + tasks: + - fail: + msg: "openshift_rolling_restart_mode must be set to either 'services' or 'system'" + when: openshift_rolling_restart_mode is defined and openshift_rolling_restart_mode not in ["services", "system"] + - openshift_facts: + role: "{{ item.role }}" + local_facts: "{{ item.local_facts }}" + with_items: + - role: common + local_facts: + rolling_restart_mode: "{{ openshift_rolling_restart_mode | default('services') }}" + - role: master + local_facts: + cluster_method: "{{ openshift_master_cluster_method | default(None) }}" + +# Creating a temp file on localhost, we then check each system that will +# be rebooted to see if that file exists, if so we know we're running +# ansible on a machine that needs a reboot, and we need to error out. +- name: Create temp file on localhost + hosts: localhost + connection: local + become: no + gather_facts: no + tasks: + - local_action: command mktemp + register: mktemp + changed_when: false + +- name: Check if temp file exists on any masters + hosts: oo_masters_to_config + tasks: + - stat: path="{{ hostvars.localhost.mktemp.stdout }}" + register: exists + changed_when: false + +- name: Cleanup temp file on localhost + hosts: localhost + connection: local + become: no + gather_facts: no + tasks: + - file: path="{{ hostvars.localhost.mktemp.stdout }}" state=absent + changed_when: false + +- name: Warn if restarting the system where ansible is running + hosts: oo_masters_to_config + tasks: + - pause: + prompt: > + Warning: Running playbook from a host that will be restarted! + Press CTRL+C and A to abort playbook execution. You may + continue by pressing ENTER but the playbook will stop + executing after this system has been restarted and services + must be verified manually. To only restart services, set + openshift_master_rolling_restart_mode=services in host + inventory and relaunch the playbook. + when: exists.stat.exists and openshift.common.rolling_restart_mode == 'system' + - set_fact: + current_host: "{{ exists.stat.exists }}" + when: openshift.common.rolling_restart_mode == 'system' + +- name: Determine which masters are currently active + hosts: oo_masters_to_config + any_errors_fatal: true + tasks: + - name: Check master service status + command: > + systemctl is-active {{ openshift.common.service_type }}-master + register: active_check_output + when: openshift.master.cluster_method | default(None) == 'pacemaker' + failed_when: false + changed_when: false + # Any master which did not report 'active' or 'inactive' is likely + # unhealthy. Other possible states are 'unknown' or 'failed'. + - fail: + msg: > + Got invalid service state from {{ openshift.common.service_type }}-master + on {{ inventory_hostname }}. Please verify pacemaker cluster. + when: openshift.master.cluster_method | default(None) == 'pacemaker' and active_check_output.stdout not in ['active', 'inactive'] + - set_fact: + is_active: "{{ active_check_output.stdout == 'active' }}" + when: openshift.master.cluster_method | default(None) == 'pacemaker' + +- name: Evaluate master groups + hosts: localhost + become: no + tasks: + - name: Evaluate oo_active_masters + add_host: + name: "{{ item }}" + groups: oo_active_masters + ansible_ssh_user: "{{ g_ssh_user | default(omit) }}" + ansible_sudo: "{{ g_sudo | default(omit) }}" + with_items: "{{ groups.oo_masters_to_config | default([]) }}" + when: (hostvars[item]['is_active'] | default(false)) | bool + - name: Evaluate oo_current_masters + add_host: + name: "{{ item }}" + groups: oo_current_masters + ansible_ssh_user: "{{ g_ssh_user | default(omit) }}" + ansible_sudo: "{{ g_sudo | default(omit) }}" + with_items: "{{ groups.oo_masters_to_config | default([]) }}" + when: (hostvars[item]['current_host'] | default(false)) | bool + +- name: Validate pacemaker cluster + hosts: oo_active_masters + tasks: + - name: Retrieve pcs status + command: pcs status + register: pcs_status_output + changed_when: false + - fail: + msg: > + Pacemaker cluster validation failed. One or more nodes are not online. + when: not (pcs_status_output.stdout | validate_pcs_cluster(groups.oo_masters_to_config)) | bool + +- name: Restart masters + hosts: oo_masters_to_config:!oo_active_masters:!oo_current_masters + vars: + openshift_master_ha: "{{ groups.oo_masters_to_config | length > 1 }}" + serial: 1 + tasks: + - include: restart_hosts.yml + when: openshift.common.rolling_restart_mode == 'system' + - include: restart_services.yml + when: openshift.common.rolling_restart_mode == 'services' + +- name: Restart active masters + hosts: oo_active_masters + serial: 1 + tasks: + - include: restart_hosts_pacemaker.yml + when: openshift.common.rolling_restart_mode == 'system' + - include: restart_services_pacemaker.yml + when: openshift.common.rolling_restart_mode == 'services' + +- name: Restart current masters + hosts: oo_current_masters + serial: 1 + tasks: + - include: restart_hosts.yml + when: openshift.common.rolling_restart_mode == 'system' + - include: restart_services.yml + when: openshift.common.rolling_restart_mode == 'services' diff --git a/playbooks/common/openshift-master/restart_hosts.yml b/playbooks/common/openshift-master/restart_hosts.yml new file mode 100644 index 000000000..ff206f5a2 --- /dev/null +++ b/playbooks/common/openshift-master/restart_hosts.yml @@ -0,0 +1,39 @@ +- name: Restart master system + # https://github.com/ansible/ansible/issues/10616 + shell: sleep 2 && shutdown -r now "OpenShift Ansible master rolling restart" + async: 1 + poll: 0 + ignore_errors: true + become: yes +# When cluster_method != pacemaker we can ensure the api_port is +# available. +- name: Wait for master API to come back online + become: no + local_action: + module: wait_for + host="{{ inventory_hostname }}" + state=started + delay=10 + port="{{ openshift.master.api_port }}" + when: openshift.master.cluster_method != 'pacemaker' +- name: Wait for master to start + become: no + local_action: + module: wait_for + host="{{ inventory_hostname }}" + state=started + delay=10 + port=22 + when: openshift.master.cluster_method == 'pacemaker' +- name: Wait for master to become available + command: pcs status + register: pcs_status_output + until: pcs_status_output.stdout | validate_pcs_cluster([inventory_hostname]) | bool + retries: 15 + delay: 2 + changed_when: false + when: openshift.master.cluster_method == 'pacemaker' +- fail: + msg: > + Pacemaker cluster validation failed {{ inventory hostname }} is not online. + when: openshift.master.cluster_method == 'pacemaker' and not (pcs_status_output.stdout | validate_pcs_cluster([inventory_hostname])) | bool diff --git a/playbooks/common/openshift-master/restart_hosts_pacemaker.yml b/playbooks/common/openshift-master/restart_hosts_pacemaker.yml new file mode 100644 index 000000000..c9219e8de --- /dev/null +++ b/playbooks/common/openshift-master/restart_hosts_pacemaker.yml @@ -0,0 +1,25 @@ +- name: Fail over master resource + command: > + pcs resource move master {{ hostvars | oo_select_keys(groups['oo_masters_to_config']) | oo_collect('openshift.common.hostname', {'is_active': 'False'}) | list | first }} +- name: Wait for master API to come back online + become: no + local_action: + module: wait_for + host="{{ openshift.master.cluster_hostname }}" + state=started + delay=10 + port="{{ openshift.master.api_port }}" +- name: Restart master system + # https://github.com/ansible/ansible/issues/10616 + shell: sleep 2 && shutdown -r now "OpenShift Ansible master rolling restart" + async: 1 + poll: 0 + ignore_errors: true + become: yes +- name: Wait for master to start + become: no + local_action: + module: wait_for + host="{{ inventory_hostname }}" + state=started + delay=10 diff --git a/playbooks/common/openshift-master/restart_services.yml b/playbooks/common/openshift-master/restart_services.yml new file mode 100644 index 000000000..5e539cd65 --- /dev/null +++ b/playbooks/common/openshift-master/restart_services.yml @@ -0,0 +1,27 @@ +- name: Restart master + service: + name: "{{ openshift.common.service_type }}-master" + state: restarted + when: not openshift_master_ha | bool +- name: Restart master API + service: + name: "{{ openshift.common.service_type }}-master-api" + state: restarted + when: openshift_master_ha | bool and openshift.master.cluster_method != 'pacemaker' +- name: Wait for master API to come back online + become: no + local_action: + module: wait_for + host="{{ inventory_hostname }}" + state=started + delay=10 + port="{{ openshift.master.api_port }}" + when: openshift_master_ha | bool and openshift.master.cluster_method != 'pacemaker' +- name: Restart master controllers + service: + name: "{{ openshift.common.service_type }}-master-controllers" + state: restarted + # Ignore errrors since it is possible that type != simple for + # pre-3.1.1 installations. + ignore_errors: true + when: openshift_master_ha | bool and openshift.master.cluster_method != 'pacemaker' diff --git a/playbooks/common/openshift-master/restart_services_pacemaker.yml b/playbooks/common/openshift-master/restart_services_pacemaker.yml new file mode 100644 index 000000000..e738f3fb6 --- /dev/null +++ b/playbooks/common/openshift-master/restart_services_pacemaker.yml @@ -0,0 +1,10 @@ +- name: Restart master services + command: pcs resource restart master +- name: Wait for master API to come back online + become: no + local_action: + module: wait_for + host="{{ openshift.master.cluster_hostname }}" + state=started + delay=10 + port="{{ openshift.master.api_port }}" diff --git a/playbooks/common/openshift-node/config.yml b/playbooks/common/openshift-node/config.yml index 483a7768c..1d31657ed 100644 --- a/playbooks/common/openshift-node/config.yml +++ b/playbooks/common/openshift-node/config.yml @@ -154,21 +154,15 @@ validate_checksum: yes with_items: nodes_needing_certs -- name: Configure node instances +- name: Deploy node certificates hosts: oo_nodes_to_config vars: sync_tmpdir: "{{ hostvars.localhost.mktemp.stdout }}" - openshift_node_master_api_url: "{{ hostvars[groups.oo_first_master.0].openshift.master.api_url }}" - # TODO: Prefix flannel role variables. - etcd_urls: "{{ hostvars[groups.oo_first_master.0].openshift.master.etcd_urls }}" - embedded_etcd: "{{ hostvars[groups.oo_first_master.0].openshift.master.embedded_etcd }}" - openshift_node_first_master_ip: "{{ hostvars[groups.oo_first_master.0].openshift.common.ip }}" - pre_tasks: + tasks: - name: Ensure certificate directory exists file: path: "{{ node_cert_dir }}" state: directory - # TODO: notify restart node # possibly test service started time against certificate/config file # timestamps in node to trigger notify @@ -177,8 +171,44 @@ src: "{{ sync_tmpdir }}/{{ node_subdir }}.tgz" dest: "{{ node_cert_dir }}" when: certs_missing + +- name: Evaluate node groups + hosts: localhost + become: no + tasks: + - name: Evaluate oo_containerized_master_nodes + add_host: + name: "{{ item }}" + groups: oo_containerized_master_nodes + ansible_ssh_user: "{{ g_ssh_user | default(omit) }}" + ansible_sudo: "{{ g_sudo | default(omit) }}" + with_items: "{{ groups.oo_nodes_to_config | default([]) }}" + when: hostvars[item].openshift.common.is_containerized | bool and (item in groups.oo_nodes_to_config and item in groups.oo_masters_to_config) + +- name: Configure node instances + hosts: oo_containerized_master_nodes + serial: 1 + vars: + openshift_node_master_api_url: "{{ hostvars[groups.oo_first_master.0].openshift.master.api_url }}" + openshift_node_first_master_ip: "{{ hostvars[groups.oo_first_master.0].openshift.common.ip }}" roles: - openshift_node + +- name: Configure node instances + hosts: oo_nodes_to_config:!oo_containerized_master_nodes + vars: + openshift_node_master_api_url: "{{ hostvars[groups.oo_first_master.0].openshift.master.api_url }}" + openshift_node_first_master_ip: "{{ hostvars[groups.oo_first_master.0].openshift.common.ip }}" + roles: + - openshift_node + +- name: Additional node config + hosts: oo_nodes_to_config + vars: + # TODO: Prefix flannel role variables. + etcd_urls: "{{ hostvars[groups.oo_first_master.0].openshift.master.etcd_urls }}" + embedded_etcd: "{{ hostvars[groups.oo_first_master.0].openshift.master.embedded_etcd }}" + roles: - role: flannel when: openshift.common.use_flannel | bool - role: nickhammond.logrotate @@ -215,6 +245,19 @@ | oo_collect('openshift.common.hostname') }}" openshift_node_vars: "{{ hostvars | oo_select_keys(groups['oo_nodes_to_config']) }}" pre_tasks: - + # Necessary because when you're on a node that's also a master the master will be + # restarted after the node restarts docker and it will take up to 60 seconds for + # systemd to start the master again + - name: Wait for master API to become available before proceeding + # Using curl here since the uri module requires python-httplib2 and + # wait_for port doesn't provide health information. + command: > + curl -k --head --silent {{ openshift.master.api_url }} + register: api_available_output + until: api_available_output.stdout.find("200 OK") != -1 + retries: 120 + delay: 1 + changed_when: false + when: openshift.common.is_containerized | bool roles: - openshift_manage_node diff --git a/playbooks/gce/openshift-cluster/config.yml b/playbooks/gce/openshift-cluster/config.yml index 3231ecc8e..84a3f84d4 100644 --- a/playbooks/gce/openshift-cluster/config.yml +++ b/playbooks/gce/openshift-cluster/config.yml @@ -13,4 +13,3 @@ openshift_debug_level: "{{ debug_level }}" openshift_deployment_type: "{{ deployment_type }}" openshift_hostname: "{{ gce_private_ip }}" - openshift_use_openshift_sdn: "{{ do_we_use_openshift_sdn }}" diff --git a/playbooks/gce/openshift-cluster/join_node.yml b/playbooks/gce/openshift-cluster/join_node.yml index acf5e5110..75343dffa 100644 --- a/playbooks/gce/openshift-cluster/join_node.yml +++ b/playbooks/gce/openshift-cluster/join_node.yml @@ -48,6 +48,4 @@ openshift_debug_level: 4 openshift_deployment_type: "{{ deployment_type }}" openshift_hostname: "{{ ansible_default_ipv4.address }}" - openshift_use_openshift_sdn: true openshift_node_labels: "{{ lookup('oo_option', 'openshift_node_labels') }} " - os_sdn_network_plugin_name: "redhat/openshift-ovs-subnet" diff --git a/playbooks/gce/openshift-cluster/vars.yml b/playbooks/gce/openshift-cluster/vars.yml index 7fb13c7a6..f004a9e6b 100644 --- a/playbooks/gce/openshift-cluster/vars.yml +++ b/playbooks/gce/openshift-cluster/vars.yml @@ -1,8 +1,5 @@ --- -do_we_use_openshift_sdn: true -sdn_network_plugin: redhat/openshift-ovs-subnet debug_level: 2 -# os_sdn_network_plugin_name can be ovssubnet or multitenant, see https://docs.openshift.org/latest/architecture/additional_concepts/sdn.html#ovssubnet-plugin-operation deployment_rhel7_ent_base: image: rhel-7 diff --git a/roles/chrony/README.md b/roles/chrony/README.md new file mode 100644 index 000000000..bf15d9669 --- /dev/null +++ b/roles/chrony/README.md @@ -0,0 +1,31 @@ +Role Name +========= + +A role to configure chrony as the ntp client + +Requirements +------------ + + +Role Variables +-------------- + +chrony_ntp_servers: a list of ntp servers to use the chrony.conf file + +Dependencies +------------ + +roles/lib_timedatectl + +Example Playbook +---------------- + +License +------- + +Apache 2.0 + +Author Information +------------------ + +Openshift Operations diff --git a/roles/chrony/defaults/main.yml b/roles/chrony/defaults/main.yml new file mode 100644 index 000000000..95576e666 --- /dev/null +++ b/roles/chrony/defaults/main.yml @@ -0,0 +1,2 @@ +--- +# defaults file for chrony diff --git a/roles/chrony/handlers/main.yml b/roles/chrony/handlers/main.yml new file mode 100644 index 000000000..1973c79e2 --- /dev/null +++ b/roles/chrony/handlers/main.yml @@ -0,0 +1,5 @@ +--- +- name: Restart chronyd + service: + name: chronyd + state: restarted diff --git a/roles/chrony/meta/main.yml b/roles/chrony/meta/main.yml new file mode 100644 index 000000000..85595d7c3 --- /dev/null +++ b/roles/chrony/meta/main.yml @@ -0,0 +1,18 @@ +--- +galaxy_info: + author: Openshift Operations + description: Configure chrony as an ntp server + company: Red Hat + license: Apache 2.0 + min_ansible_version: 1.9.2 + platforms: + - name: EL + versions: + - 7 + - name: Fedora + versions: + - all + categories: + - system +dependencies: +- roles/lib_timedatectl diff --git a/roles/chrony/tasks/main.yml b/roles/chrony/tasks/main.yml new file mode 100644 index 000000000..fae6d8e4c --- /dev/null +++ b/roles/chrony/tasks/main.yml @@ -0,0 +1,30 @@ +--- +- name: remove ntp package + yum: + name: ntp + state: absent + +- name: ensure chrony package is installed + yum: + name: chrony + state: installed + +- name: Install /etc/chrony.conf + template: + src: chrony.conf.j2 + dest: /etc/chrony.conf + owner: root + group: root + mode: 0644 + notify: + - Restart chronyd + +- name: enabled timedatectl set-ntp yes + timedatectl: + ntp: True + +- name: + service: + name: chronyd + state: started + enabled: yes diff --git a/roles/chrony/templates/chrony.conf.j2 b/roles/chrony/templates/chrony.conf.j2 new file mode 100644 index 000000000..de43b6364 --- /dev/null +++ b/roles/chrony/templates/chrony.conf.j2 @@ -0,0 +1,45 @@ +# Use public servers from the pool.ntp.org project. +# Please consider joining the pool (http://www.pool.ntp.org/join.html). +{% for server in chrony_ntp_servers %} +server {{ server }} iburst +{% endfor %} + +# Ignore stratum in source selection. +stratumweight 0 + +# Record the rate at which the system clock gains/losses time. +driftfile /var/lib/chrony/drift + +# Enable kernel RTC synchronization. +rtcsync + +# In first three updates step the system clock instead of slew +# if the adjustment is larger than 10 seconds. +makestep 10 3 + +# Allow NTP client access from local network. +#allow 192.168/16 + +# Listen for commands only on localhost. +bindcmdaddress 127.0.0.1 +bindcmdaddress ::1 + +# Serve time even if not synchronized to any NTP server. +#local stratum 10 + +keyfile /etc/chrony.keys + +# Specify the key used as password for chronyc. +commandkey 1 + +# Generate command key if missing. +generatecommandkey + +# Disable logging of client accesses. +noclientlog + +# Send a message to syslog if a clock adjustment is larger than 0.5 seconds. +logchange 0.5 + +logdir /var/log/chrony +#log measurements statistics tracking diff --git a/roles/chrony/vars/main.yml b/roles/chrony/vars/main.yml new file mode 100644 index 000000000..061a21547 --- /dev/null +++ b/roles/chrony/vars/main.yml @@ -0,0 +1,2 @@ +--- +# vars file for chrony diff --git a/roles/haproxy/handlers/main.yml b/roles/haproxy/handlers/main.yml index ee60adcab..5b8691b26 100644 --- a/roles/haproxy/handlers/main.yml +++ b/roles/haproxy/handlers/main.yml @@ -3,3 +3,4 @@ service: name: haproxy state: restarted + when: not (haproxy_start_result_changed | default(false) | bool) diff --git a/roles/haproxy/tasks/main.yml b/roles/haproxy/tasks/main.yml index 97f870829..0b8370ce2 100644 --- a/roles/haproxy/tasks/main.yml +++ b/roles/haproxy/tasks/main.yml @@ -19,6 +19,5 @@ enabled: yes register: start_result -- name: Pause 30 seconds if haproxy was just started - pause: seconds=30 - when: start_result | changed +- set_fact: + haproxy_start_result_changed: "{{ start_result | changed }}" diff --git a/roles/lib_timedatectl/library/timedatectl.py b/roles/lib_timedatectl/library/timedatectl.py new file mode 100644 index 000000000..b6eab5918 --- /dev/null +++ b/roles/lib_timedatectl/library/timedatectl.py @@ -0,0 +1,74 @@ +#!/usr/bin/env python +''' + timedatectl ansible module + + This module supports setting ntp enabled +''' +import subprocess + + + + +def do_timedatectl(options=None): + ''' subprocess timedatectl ''' + + cmd = ['/usr/bin/timedatectl'] + if options: + cmd += options.split() + + proc = subprocess.Popen(cmd, stdin=None, stdout=subprocess.PIPE) + proc.wait() + return proc.stdout.read() + +def main(): + ''' Ansible module for timedatectl + ''' + + module = AnsibleModule( + argument_spec=dict( + #state=dict(default='enabled', type='str'), + ntp=dict(default=True, type='bool'), + ), + #supports_check_mode=True + ) + + # do something + ntp_enabled = False + + results = do_timedatectl() + + for line in results.split('\n'): + if 'NTP enabled' in line: + if 'yes' in line: + ntp_enabled = True + + ######## + # Enable NTP + ######## + if module.params['ntp']: + if ntp_enabled: + module.exit_json(changed=False, results="enabled", state="enabled") + + # Enable it + # Commands to enable ntp + else: + results = do_timedatectl('set-ntp yes') + module.exit_json(changed=True, results="enabled", state="enabled", cmdout=results) + + ######### + # Disable NTP + ######### + else: + if not ntp_enabled: + module.exit_json(changed=False, results="disabled", state="disabled") + + results = do_timedatectl('set-ntp no') + module.exit_json(changed=True, results="disabled", state="disabled") + + module.exit_json(failed=True, changed=False, results="Something went wrong", state="unknown") + +# Pylint is getting in the way of basic Ansible +# pylint: disable=redefined-builtin,wildcard-import,unused-wildcard-import +from ansible.module_utils.basic import * + +main() diff --git a/roles/lib_zabbix/library/zbx_action.py b/roles/lib_zabbix/library/zbx_action.py index c08bef4f7..2f9524556 100644 --- a/roles/lib_zabbix/library/zbx_action.py +++ b/roles/lib_zabbix/library/zbx_action.py @@ -81,6 +81,61 @@ def filter_differences(zabbix_filters, user_filters): return rval +def opconditions_diff(zab_val, user_val): + ''' Report whether there are differences between opconditions on + zabbix and opconditions supplied by user ''' + + if len(zab_val) != len(user_val): + return True + + for z_cond, u_cond in zip(zab_val, user_val): + if not all([str(u_cond[op_key]) == z_cond[op_key] for op_key in \ + ['conditiontype', 'operator', 'value']]): + return True + + return False + +def opmessage_diff(zab_val, user_val): + ''' Report whether there are differences between opmessage on + zabbix and opmessage supplied by user ''' + + for op_msg_key, op_msg_val in user_val.items(): + if zab_val[op_msg_key] != str(op_msg_val): + return True + + return False + +def opmessage_grp_diff(zab_val, user_val): + ''' Report whether there are differences between opmessage_grp + on zabbix and opmessage_grp supplied by user ''' + + zab_grp_ids = set([ugrp['usrgrpid'] for ugrp in zab_val]) + usr_grp_ids = set([ugrp['usrgrpid'] for ugrp in user_val]) + if usr_grp_ids != zab_grp_ids: + return True + + return False + +def opmessage_usr_diff(zab_val, user_val): + ''' Report whether there are differences between opmessage_usr + on zabbix and opmessage_usr supplied by user ''' + + zab_usr_ids = set([usr['usrid'] for usr in zab_val]) + usr_ids = set([usr['usrid'] for usr in user_val]) + if usr_ids != zab_usr_ids: + return True + + return False + +def opcommand_diff(zab_op_cmd, usr_op_cmd): + ''' Check whether user-provided opcommand matches what's already + stored in Zabbix ''' + + for usr_op_cmd_key, usr_op_cmd_val in usr_op_cmd.items(): + if zab_op_cmd[usr_op_cmd_key] != str(usr_op_cmd_val): + return True + return False + def host_in_zabbix(zab_hosts, usr_host): ''' Check whether a particular user host is already in the Zabbix list of hosts ''' @@ -106,23 +161,11 @@ def hostlist_in_zabbix(zab_hosts, usr_hosts): return True -def opcommand_diff(zab_op_cmd, usr_op_cmd): - ''' Check whether user-provided opcommand matches what's already - stored in Zabbix ''' - - for usr_op_cmd_key, usr_op_cmd_val in usr_op_cmd.items(): - if zab_op_cmd[usr_op_cmd_key] != str(usr_op_cmd_val): - return True - return False - -# This logic is quite complex. We are comparing two lists of dictionaries. -# The outer for-loops allow us to descend down into both lists at the same time -# and then walk over the key,val pairs of the incoming user dict's changes -# or updates. The if-statements are looking at different sub-object types and -# comparing them. The other suggestion on how to write this is to write a recursive -# compare function but for the time constraints and for complexity I decided to go -# this route. -# pylint: disable=too-many-branches +# We are comparing two lists of dictionaries (the one stored on zabbix and the +# one the user is providing). For each type of operation, determine whether there +# is a difference between what is stored on zabbix and what the user is providing. +# If there is a difference, we take the user-provided data for what needs to +# be stored/updated into zabbix. def operation_differences(zabbix_ops, user_ops): '''Determine the differences from user and zabbix for operations''' @@ -132,49 +175,41 @@ def operation_differences(zabbix_ops, user_ops): rval = {} for zab, user in zip(zabbix_ops, user_ops): - for key, val in user.items(): - if key == 'opconditions': - if len(zab[key]) != len(val): - rval[key] = val - break - for z_cond, u_cond in zip(zab[key], user[key]): - if not all([str(u_cond[op_key]) == z_cond[op_key] for op_key in \ - ['conditiontype', 'operator', 'value']]): - rval[key] = val - break - elif key == 'opmessage': - # Verify each passed param matches - for op_msg_key, op_msg_val in val.items(): - if zab[key][op_msg_key] != str(op_msg_val): - rval[key] = val - break - - elif key == 'opmessage_grp': - zab_grp_ids = set([ugrp['usrgrpid'] for ugrp in zab[key]]) - usr_grp_ids = set([ugrp['usrgrpid'] for ugrp in val]) - if usr_grp_ids != zab_grp_ids: - rval[key] = val - - elif key == 'opmessage_usr': - zab_usr_ids = set([usr['userid'] for usr in zab[key]]) - usr_ids = set([usr['userid'] for usr in val]) - if usr_ids != zab_usr_ids: - rval[key] = val - - elif key == 'opcommand': - if opcommand_diff(zab[key], val): - rval[key] = val - break + for oper in user.keys(): + if oper == 'opconditions' and opconditions_diff(zab[oper], \ + user[oper]): + rval[oper] = user[oper] + + elif oper == 'opmessage' and opmessage_diff(zab[oper], \ + user[oper]): + rval[oper] = user[oper] + + elif oper == 'opmessage_grp' and opmessage_grp_diff(zab[oper], \ + user[oper]): + rval[oper] = user[oper] + + elif oper == 'opmessage_usr' and opmessage_usr_diff(zab[oper], \ + user[oper]): + rval[oper] = user[oper] + + elif oper == 'opcommand' and opcommand_diff(zab[oper], \ + user[oper]): + rval[oper] = user[oper] # opcommand_grp can be treated just like opcommand_hst # as opcommand_grp[] is just a list of groups - elif key == 'opcommand_hst' or key == 'opcommand_grp': - if not hostlist_in_zabbix(zab[key], val): - rval[key] = val - break + elif oper == 'opcommand_hst' or oper == 'opcommand_grp': + if not hostlist_in_zabbix(zab[oper], user[oper]): + rval[oper] = user[oper] + + # if it's any other type of operation than the ones tested above + # just do a direct compare + elif oper not in ['opconditions', 'opmessage', 'opmessage_grp', + 'opmessage_usr', 'opcommand', 'opcommand_hst', + 'opcommand_grp'] \ + and str(zab[oper]) != str(user[oper]): + rval[oper] = user[oper] - elif zab[key] != str(val): - rval[key] = val return rval def get_users(zapi, users): diff --git a/roles/lib_zabbix/tasks/create_template.yml b/roles/lib_zabbix/tasks/create_template.yml index 47749389e..61344357a 100644 --- a/roles/lib_zabbix/tasks/create_template.yml +++ b/roles/lib_zabbix/tasks/create_template.yml @@ -57,6 +57,7 @@ expression: "{{ item.expression }}" priority: "{{ item.priority }}" url: "{{ item.url | default(None, True) }}" + status: "{{ item.status | default('', True) }}" with_items: template.ztriggers when: template.ztriggers is defined diff --git a/roles/openshift_facts/library/openshift_facts.py b/roles/openshift_facts/library/openshift_facts.py index 9096f7ebf..0820989af 100755 --- a/roles/openshift_facts/library/openshift_facts.py +++ b/roles/openshift_facts/library/openshift_facts.py @@ -188,9 +188,6 @@ def normalize_gce_facts(metadata, facts): _, _, zone = metadata['instance']['zone'].rpartition('/') facts['zone'] = zone - # Default to no sdn for GCE deployments - facts['use_openshift_sdn'] = False - # GCE currently only supports a single interface facts['network']['ip'] = facts['network']['interfaces'][0]['ips'][0] pub_ip = facts['network']['interfaces'][0]['public_ips'][0] @@ -900,10 +897,6 @@ def apply_provider_facts(facts, provider_facts): if not provider_facts: return facts - use_openshift_sdn = provider_facts.get('use_openshift_sdn') - if isinstance(use_openshift_sdn, bool): - facts['common']['use_openshift_sdn'] = use_openshift_sdn - common_vars = [('hostname', 'ip'), ('public_hostname', 'public_ip')] for h_var, ip_var in common_vars: ip_value = provider_facts['network'].get(ip_var) @@ -1094,7 +1087,7 @@ class OpenShiftFacts(object): Raises: OpenShiftFactsUnsupportedRoleError: """ - known_roles = ['common', 'master', 'node', 'master_sdn', 'node_sdn', 'etcd', 'nfs'] + known_roles = ['common', 'master', 'node', 'etcd', 'nfs'] def __init__(self, role, filename, local_facts, additive_facts_to_overwrite=False): self.changed = False diff --git a/roles/openshift_facts/tasks/main.yml b/roles/openshift_facts/tasks/main.yml index 87fa99a3b..e40a1b329 100644 --- a/roles/openshift_facts/tasks/main.yml +++ b/roles/openshift_facts/tasks/main.yml @@ -10,6 +10,7 @@ shell: ls /run/ostree-booted ignore_errors: yes failed_when: false + changed_when: false register: ostree_output # Locally setup containerized facts for now diff --git a/roles/openshift_master/handlers/main.yml b/roles/openshift_master/handlers/main.yml index e1b95eda4..6b9992eea 100644 --- a/roles/openshift_master/handlers/main.yml +++ b/roles/openshift_master/handlers/main.yml @@ -2,11 +2,24 @@ - name: restart master service: name={{ openshift.common.service_type }}-master state=restarted when: (not openshift_master_ha | bool) and (not (master_service_status_changed | default(false) | bool)) + notify: Verify API Server - name: restart master api service: name={{ openshift.common.service_type }}-master-api state=restarted when: (openshift_master_ha | bool) and (not (master_api_service_status_changed | default(false) | bool)) and openshift.master.cluster_method == 'native' + notify: Verify API Server - name: restart master controllers service: name={{ openshift.common.service_type }}-master-controllers state=restarted when: (openshift_master_ha | bool) and (not (master_controllers_service_status_changed | default(false) | bool)) and openshift.master.cluster_method == 'native' + +- name: Verify API Server + # Using curl here since the uri module requires python-httplib2 and + # wait_for port doesn't provide health information. + command: > + curl -k --head --silent {{ openshift.master.api_url }} + register: api_available_output + until: api_available_output.stdout.find("200 OK") != -1 + retries: 120 + delay: 1 + changed_when: false diff --git a/roles/openshift_master/tasks/main.yml b/roles/openshift_master/tasks/main.yml index 462a7ab58..44e7e3178 100644 --- a/roles/openshift_master/tasks/main.yml +++ b/roles/openshift_master/tasks/main.yml @@ -269,6 +269,7 @@ service: name={{ openshift.common.service_type }}-master enabled=yes state=started when: not openshift_master_ha | bool register: start_result + notify: Verify API Server - name: Stop and disable non HA master when running HA service: name={{ openshift.common.service_type }}-master enabled=no state=stopped @@ -287,6 +288,20 @@ master_api_service_status_changed: "{{ start_result | changed }}" when: openshift_master_ha | bool and openshift.master.cluster_method == 'native' +# A separate wait is required here for native HA since notifies will +# be resolved after all tasks in the role. +- name: Wait for API to become available + # Using curl here since the uri module requires python-httplib2 and + # wait_for port doesn't provide health information. + command: > + curl -k --head --silent {{ openshift.master.api_url }} + register: api_available_output + until: api_available_output.stdout.find("200 OK") != -1 + retries: 120 + delay: 1 + changed_when: false + when: openshift_master_ha | bool and openshift.master.cluster_method == 'native' and master_api_service_status_changed | bool + - name: Start and enable master controller service: name={{ openshift.common.service_type }}-master-controllers enabled=yes state=started when: openshift_master_ha | bool and openshift.master.cluster_method == 'native' diff --git a/roles/openshift_node/tasks/main.yml b/roles/openshift_node/tasks/main.yml index 0828d8e2c..9035248f9 100644 --- a/roles/openshift_node/tasks/main.yml +++ b/roles/openshift_node/tasks/main.yml @@ -103,6 +103,21 @@ - name: Additional storage plugin configuration include: storage_plugins/main.yml +# Necessary because when you're on a node that's also a master the master will be +# restarted after the node restarts docker and it will take up to 60 seconds for +# systemd to start the master again +- name: Wait for master API to become available before proceeding + # Using curl here since the uri module requires python-httplib2 and + # wait_for port doesn't provide health information. + command: > + curl -k --head --silent {{ openshift_node_master_api_url }} + register: api_available_output + until: api_available_output.stdout.find("200 OK") != -1 + retries: 120 + delay: 1 + changed_when: false + when: openshift.common.is_containerized | bool + - name: Start and enable node service: name={{ openshift.common.service_type }}-node enabled=yes state=started register: start_result diff --git a/roles/openshift_node/tasks/storage_plugins/nfs.yml b/roles/openshift_node/tasks/storage_plugins/nfs.yml index 1edf21d9b..14a613786 100644 --- a/roles/openshift_node/tasks/storage_plugins/nfs.yml +++ b/roles/openshift_node/tasks/storage_plugins/nfs.yml @@ -1,4 +1,8 @@ --- +- name: Install NFS storage plugin dependencies + action: "{{ ansible_pkg_mgr }} name=nfs-utils state=present" + when: not openshift.common.is_atomic | bool + - name: Set seboolean to allow nfs storage plugin access from containers seboolean: name: virt_use_nfs diff --git a/roles/os_firewall/defaults/main.yml b/roles/os_firewall/defaults/main.yml index bcf1d9a34..e3176e611 100644 --- a/roles/os_firewall/defaults/main.yml +++ b/roles/os_firewall/defaults/main.yml @@ -1,2 +1,3 @@ --- +os_firewall_enabled: True os_firewall_use_firewalld: True diff --git a/roles/os_firewall/tasks/main.yml b/roles/os_firewall/tasks/main.yml index ad89ef97c..076e5e311 100644 --- a/roles/os_firewall/tasks/main.yml +++ b/roles/os_firewall/tasks/main.yml @@ -1,6 +1,6 @@ --- - include: firewall/firewalld.yml - when: os_firewall_use_firewalld + when: os_firewall_enabled | bool and os_firewall_use_firewalld | bool - include: firewall/iptables.yml - when: not os_firewall_use_firewalld + when: os_firewall_enabled | bool and not os_firewall_use_firewalld | bool diff --git a/roles/os_zabbix/tasks/main.yml b/roles/os_zabbix/tasks/main.yml index a8b65dd56..7552086d4 100644 --- a/roles/os_zabbix/tasks/main.yml +++ b/roles/os_zabbix/tasks/main.yml @@ -1,8 +1,4 @@ --- -- fail: - msg: "Zabbix config is not yet supported on atomic hosts" - when: openshift.common.is_containerized | bool - - name: Main List all templates zbx_template: zbx_server: "{{ ozb_server }}" diff --git a/roles/os_zabbix/vars/template_docker.yml b/roles/os_zabbix/vars/template_docker.yml index a05e552e3..dd13e76f7 100644 --- a/roles/os_zabbix/vars/template_docker.yml +++ b/roles/os_zabbix/vars/template_docker.yml @@ -72,10 +72,12 @@ g_template_docker: url: 'https://github.com/openshift/ops-sop/blob/master/V3/Alerts/check_docker_ping.asciidoc' priority: high + # Re-enable for OpenShift 3.1.1 (https://bugzilla.redhat.com/show_bug.cgi?id=1292971#c6) - name: 'docker.container.dns.resolution failed on {HOST.NAME}' expression: '{Template Docker:docker.container.dns.resolution.min(#3)}>0' url: 'https://github.com/openshift/ops-sop/blob/master/V3/Alerts/check_docker_dns.asciidoc' priority: average + status: disabled - name: 'docker.container.existing.dns.resolution.failed on {HOST.NAME}' expression: '{Template Docker:docker.container.existing.dns.resolution.failed.min(#3)}>0' diff --git a/roles/oso_monitoring_tools/README.md b/roles/oso_monitoring_tools/README.md new file mode 100644 index 000000000..4215f9eeb --- /dev/null +++ b/roles/oso_monitoring_tools/README.md @@ -0,0 +1,54 @@ +Role Name +========= + +This role will install the Openshift Monitoring Utilities + +Requirements +------------ + +Any pre-requisites that may not be covered by Ansible itself or the role should be mentioned here. For instance, if the role uses the EC2 module, it may be a good idea to mention in this section that the boto package is required. + +Role Variables +-------------- + +osomt_zagg_client_config + +from vars/main.yml: + +osomt_zagg_client_config: + host: + name: "{{ osomt_host_name }}" + zagg: + url: "{{ osomt_zagg_url }}" + user: "{{ osomt_zagg_user }}" + pass: "{{ osomt_zagg_password }}" + ssl_verify: "{{ osomt_zagg_ssl_verify }}" + verbose: "{{ osomt_zagg_verbose }}" + debug: "{{ osomt_zagg_debug }}" + +Dependencies +------------ + +None + +Example Playbook +---------------- + +- role: "oso_monitoring_tools" + osomt_host_name: hostname + osomt_zagg_url: http://path.to/zagg_web + osomt_zagg_user: admin + osomt_zagg_password: password + osomt_zagg_ssl_verify: True + osomt_zagg_verbose: False + osomt_zagg_debug: False + +License +------- + +BSD + +Author Information +------------------ + +Openshift Operations diff --git a/roles/oso_monitoring_tools/defaults/main.yml b/roles/oso_monitoring_tools/defaults/main.yml new file mode 100644 index 000000000..a17424f25 --- /dev/null +++ b/roles/oso_monitoring_tools/defaults/main.yml @@ -0,0 +1,2 @@ +--- +# defaults file for oso_monitoring_tools diff --git a/roles/oso_monitoring_tools/handlers/main.yml b/roles/oso_monitoring_tools/handlers/main.yml new file mode 100644 index 000000000..cefa780ab --- /dev/null +++ b/roles/oso_monitoring_tools/handlers/main.yml @@ -0,0 +1,2 @@ +--- +# handlers file for oso_monitoring_tools diff --git a/roles/oso_monitoring_tools/meta/main.yml b/roles/oso_monitoring_tools/meta/main.yml new file mode 100644 index 000000000..9c42b68dc --- /dev/null +++ b/roles/oso_monitoring_tools/meta/main.yml @@ -0,0 +1,8 @@ +--- +galaxy_info: + author: OpenShift Operations + description: Install Openshift Monitoring tools + company: Red Hat, Inc + license: ASL 2.0 + min_ansible_version: 1.2 +dependencies: [] diff --git a/roles/oso_monitoring_tools/tasks/main.yml b/roles/oso_monitoring_tools/tasks/main.yml new file mode 100644 index 000000000..c90fc56e2 --- /dev/null +++ b/roles/oso_monitoring_tools/tasks/main.yml @@ -0,0 +1,18 @@ +--- +# tasks file for oso_monitoring_tools +- name: Install the Openshift Tools RPMS + yum: + name: "{{ item }}" + state: latest + with_items: + - openshift-tools-scripts-monitoring-zagg-client + - python-openshift-tools-monitoring-zagg + - python-openshift-tools-monitoring-zabbix + +- debug: var=g_zagg_client_config + +- name: Generate the /etc/openshift_tools/zagg_client.yaml config file + copy: + content: "{{ osomt_zagg_client_config | to_nice_yaml }}" + dest: /etc/openshift_tools/zagg_client.yaml + mode: "644" diff --git a/roles/oso_monitoring_tools/vars/main.yml b/roles/oso_monitoring_tools/vars/main.yml new file mode 100644 index 000000000..3538ba30b --- /dev/null +++ b/roles/oso_monitoring_tools/vars/main.yml @@ -0,0 +1,12 @@ +--- +# vars file for oso_monitoring_tools +osomt_zagg_client_config: + host: + name: "{{ osomt_host_name }}" + zagg: + url: "{{ osomt_zagg_url }}" + user: "{{ osomt_zagg_user }}" + pass: "{{ osomt_zagg_password }}" + ssl_verify: "{{ osomt_zagg_ssl_verify }}" + verbose: "{{ osomt_zagg_verbose }}" + debug: "{{ osomt_zagg_debug }}" diff --git a/utils/src/ooinstall/cli_installer.py b/utils/src/ooinstall/cli_installer.py index 2b6c9deee..1aacf3a4b 100644 --- a/utils/src/ooinstall/cli_installer.py +++ b/utils/src/ooinstall/cli_installer.py @@ -33,9 +33,7 @@ def is_valid_hostname(hostname): def validate_prompt_hostname(hostname): if '' == hostname or is_valid_hostname(hostname): return hostname - raise click.BadParameter('"{}" appears to be an invalid hostname. ' \ - 'Please double-check this value i' \ - 'and re-enter it.'.format(hostname)) + raise click.BadParameter('Invalid hostname. Please double-check this value and re-enter it.') def get_ansible_ssh_user(): click.clear() @@ -129,14 +127,13 @@ http://docs.openshift.com/enterprise/latest/architecture/infrastructure_componen masters_set = True host_props['node'] = True - #TODO: Reenable this option once container installs are out of tech preview - rpm_or_container = click.prompt('Will this host be RPM or Container based (rpm/container)?', - type=click.Choice(['rpm', 'container']), - default='rpm') - if rpm_or_container == 'container': - host_props['containerized'] = True - else: - host_props['containerized'] = False + host_props['containerized'] = False + if oo_cfg.settings['variant_version'] != '3.0': + rpm_or_container = click.prompt('Will this host be RPM or Container based (rpm/container)?', + type=click.Choice(['rpm', 'container']), + default='rpm') + if rpm_or_container == 'container': + host_props['containerized'] = True if existing_env: host_props['new_host'] = True diff --git a/utils/src/ooinstall/openshift_ansible.py b/utils/src/ooinstall/openshift_ansible.py index 20401f812..c0d115fdc 100644 --- a/utils/src/ooinstall/openshift_ansible.py +++ b/utils/src/ooinstall/openshift_ansible.py @@ -128,6 +128,8 @@ def write_host(host, inventory, schedulable=None): facts += ' openshift_hostname={}'.format(host.hostname) if host.public_hostname: facts += ' openshift_public_hostname={}'.format(host.public_hostname) + if host.containerized: + facts += ' containerized={}'.format(host.containerized) # TODO: For not write_host is handles both master and nodes. # Technically only nodes will ever need this. diff --git a/utils/test/cli_installer_tests.py b/utils/test/cli_installer_tests.py index ea380d565..72e8521d0 100644 --- a/utils/test/cli_installer_tests.py +++ b/utils/test/cli_installer_tests.py @@ -681,9 +681,9 @@ class AttendedCliTests(OOCliFixture): run_playbook_mock.return_value = 0 cli_input = build_input(hosts=[ - ('10.0.0.1', True), - ('10.0.0.2', False), - ('10.0.0.3', False)], + ('10.0.0.1', True, False), + ('10.0.0.2', False, False), + ('10.0.0.3', False, False)], ssh_user='root', variant_num=1, confirm_facts='y') @@ -722,10 +722,10 @@ class AttendedCliTests(OOCliFixture): run_playbook_mock.return_value = 0 cli_input = build_input(hosts=[ - ('10.0.0.1', True), - ('10.0.0.2', False), + ('10.0.0.1', True, False), + ('10.0.0.2', False, False), ], - add_nodes=[('10.0.0.3', False)], + add_nodes=[('10.0.0.3', False, False)], ssh_user='root', variant_num=1, confirm_facts='y') @@ -773,9 +773,9 @@ class AttendedCliTests(OOCliFixture): mock_facts['10.0.0.2']['common']['version'] = "3.0.0" cli_input = build_input(hosts=[ - ('10.0.0.1', True), + ('10.0.0.1', True, False), ], - add_nodes=[('10.0.0.2', False)], + add_nodes=[('10.0.0.2', False, False)], ssh_user='root', variant_num=1, schedulable_masters_ok=True, @@ -796,10 +796,10 @@ class AttendedCliTests(OOCliFixture): run_playbook_mock.return_value = 0 cli_input = build_input(hosts=[ - ('10.0.0.1', True), - ('10.0.0.2', True), - ('10.0.0.3', True), - ('10.0.0.4', False)], + ('10.0.0.1', True, False), + ('10.0.0.2', True, False), + ('10.0.0.3', True, False), + ('10.0.0.4', False, False)], ssh_user='root', variant_num=1, confirm_facts='y', @@ -837,9 +837,9 @@ class AttendedCliTests(OOCliFixture): run_playbook_mock.return_value = 0 cli_input = build_input(hosts=[ - ('10.0.0.1', True), - ('10.0.0.2', True), - ('10.0.0.3', True)], + ('10.0.0.1', True, False), + ('10.0.0.2', True, False), + ('10.0.0.3', True, False)], ssh_user='root', variant_num=1, confirm_facts='y', @@ -872,10 +872,10 @@ class AttendedCliTests(OOCliFixture): run_playbook_mock.return_value = 0 cli_input = build_input(hosts=[ - ('10.0.0.1', True), - ('10.0.0.2', True), - ('10.0.0.3', False), - ('10.0.0.4', True)], + ('10.0.0.1', True, False), + ('10.0.0.2', True, False), + ('10.0.0.3', False, False), + ('10.0.0.4', True, False)], ssh_user='root', variant_num=1, confirm_facts='y', @@ -893,7 +893,7 @@ class AttendedCliTests(OOCliFixture): run_playbook_mock.return_value = 0 cli_input = build_input(hosts=[ - ('10.0.0.1', True)], + ('10.0.0.1', True, False)], ssh_user='root', variant_num=1, confirm_facts='y') @@ -921,7 +921,7 @@ class AttendedCliTests(OOCliFixture): run_playbook_mock.return_value = 0 cli_input = build_input(hosts=[ - ('10.0.0.1', True)], + ('10.0.0.1', True, False)], ssh_user='root', variant_num=2, confirm_facts='y') diff --git a/utils/test/fixture.py b/utils/test/fixture.py index 90bd9e1ef..be759578a 100644 --- a/utils/test/fixture.py +++ b/utils/test/fixture.py @@ -138,7 +138,7 @@ class OOCliFixture(OOInstallFixture): self.assertEquals(exp_hosts_to_run_on_len, len(hosts_to_run_on)) -#pylint: disable=too-many-arguments,too-many-branches +#pylint: disable=too-many-arguments,too-many-branches,too-many-statements def build_input(ssh_user=None, hosts=None, variant_num=None, add_nodes=None, confirm_facts=None, schedulable_masters_ok=None, master_lb=None): @@ -163,13 +163,19 @@ def build_input(ssh_user=None, hosts=None, variant_num=None, num_masters = 0 if hosts: i = 0 - for (host, is_master) in hosts: + for (host, is_master, is_containerized) in hosts: inputs.append(host) if is_master: inputs.append('y') num_masters += 1 else: inputs.append('n') + + if is_containerized: + inputs.append('container') + else: + inputs.append('rpm') + #inputs.append('rpm') # We should not be prompted to add more hosts if we're currently at # 2 masters, this is an invalid HA configuration, so this question @@ -196,8 +202,12 @@ def build_input(ssh_user=None, hosts=None, variant_num=None, inputs.append('y') inputs.append('1') # Add more nodes i = 0 - for (host, is_master) in add_nodes: + for (host, is_master, is_containerized) in add_nodes: inputs.append(host) + if is_containerized: + inputs.append('container') + else: + inputs.append('rpm') #inputs.append('rpm') if i < len(add_nodes) - 1: inputs.append('y') # Add more hosts |