diff options
31 files changed, 319 insertions, 106 deletions
diff --git a/.redhat-ci.inventory b/.redhat-ci.inventory index 3c8296055..23bc9923c 100644 --- a/.redhat-ci.inventory +++ b/.redhat-ci.inventory @@ -9,6 +9,7 @@ ansible_python_interpreter=/usr/bin/python3 deployment_type=origin openshift_image_tag="{{ lookup('env', 'OPENSHIFT_IMAGE_TAG') }}" openshift_master_default_subdomain="{{ lookup('env', 'RHCI_ocp_node1_IP') }}.xip.io" +openshift_check_min_host_memory_gb=1.9 [masters] ocp-master diff --git a/.tito/packages/openshift-ansible b/.tito/packages/openshift-ansible index 4e3183de1..9a340c066 100644 --- a/.tito/packages/openshift-ansible +++ b/.tito/packages/openshift-ansible @@ -1 +1 @@ -3.6.80-1 ./ +3.6.86-1 ./ diff --git a/inventory/byo/hosts.origin.example b/inventory/byo/hosts.origin.example index d52036930..5d21f8e14 100644 --- a/inventory/byo/hosts.origin.example +++ b/inventory/byo/hosts.origin.example @@ -127,6 +127,10 @@ openshift_release=v3.6 # Alternate image format string, useful if you've got your own registry mirror +# Configure this setting just on node or master +#oreg_url_master=example.com/openshift3/ose-${component}:${version} +#oreg_url_node=example.com/openshift3/ose-${component}:${version} +# For setting the configuration globally #oreg_url=example.com/openshift3/ose-${component}:${version} # If oreg_url points to a registry other than registry.access.redhat.com we can # modify image streams to point at that registry by setting the following to true diff --git a/inventory/byo/hosts.ose.example b/inventory/byo/hosts.ose.example index 647dd3d9a..662df8104 100644 --- a/inventory/byo/hosts.ose.example +++ b/inventory/byo/hosts.ose.example @@ -125,8 +125,11 @@ openshift_release=v3.6 # Tasks to run after each master is upgraded and system/services have been restarted. # openshift_master_upgrade_post_hook=/usr/share/custom/post_master.yml - # Alternate image format string, useful if you've got your own registry mirror +# Configure this setting just on node or master +#oreg_url_master=example.com/openshift3/ose-${component}:${version} +#oreg_url_node=example.com/openshift3/ose-${component}:${version} +# For setting the configuration globally #oreg_url=example.com/openshift3/ose-${component}:${version} # If oreg_url points to a registry other than registry.access.redhat.com we can # modify image streams to point at that registry by setting the following to true diff --git a/openshift-ansible.spec b/openshift-ansible.spec index 3e967a8e8..813cba0d7 100644 --- a/openshift-ansible.spec +++ b/openshift-ansible.spec @@ -9,7 +9,7 @@ %global __requires_exclude ^/usr/bin/ansible-playbook$ Name: openshift-ansible -Version: 3.6.80 +Version: 3.6.86 Release: 1%{?dist} Summary: Openshift and Atomic Enterprise Ansible License: ASL 2.0 @@ -280,6 +280,50 @@ Atomic OpenShift Utilities includes %changelog +* Tue May 30 2017 Jenkins CD Merge Bot <tdawson@redhat.com> 3.6.86-1 +- Reduce memory requirement to 2gb for fedora ci jobs (sdodson@redhat.com) +- openshift_logging: increasing *_elasticsearch_* default CPU and memory + (jwozniak@redhat.com) +- Updating python-passlib assert (ewolinet@redhat.com) +- allow to configure oreg_url specifically for node or master. refs #4233 + (tobias@tobru.ch) +- Updating registry-console version to be v3.6 instead of 3.6 + (ewolinet@redhat.com) + +* Thu May 25 2017 Jenkins CD Merge Bot <tdawson@redhat.com> 3.6.85-1 +- Prepending v to registry-console version (ewolinet@redhat.com) +- memory health check: adjust threshold for etcd (lmeyer@redhat.com) +- health checks: specify check skip reason (lmeyer@redhat.com) +- health checks: configure failure output in playbooks (lmeyer@redhat.com) +- disk/memory checks: make threshold configurable (lmeyer@redhat.com) +- Show help on how to disable checks after failure (rhcarvalho@gmail.com) +- Allow disabling checks via Ansible variable (rhcarvalho@gmail.com) +- Verify memory and disk requirements before install (rhcarvalho@gmail.com) +- filter_plugins: Allow for multiple pairs in map_from_pairs() + (jarrpa@redhat.com) + +* Wed May 24 2017 Jenkins CD Merge Bot <tdawson@redhat.com> 3.6.84-1 +- oc_process: Better error output on failed template() call (jarrpa@redhat.com) + +* Wed May 24 2017 Jenkins CD Merge Bot <tdawson@redhat.com> 3.6.83-1 +- Allow a hostname to resolve to 127.0.0.1 during validation (dms@redhat.com) + +* Wed May 24 2017 Jenkins CD Merge Bot <tdawson@redhat.com> 3.6.82-1 +- Fixing tux warnings and some final clean up (ewolinet@redhat.com) +- Appease travis (sdodson@redhat.com) +- preflight int tests: fix test flake (lmeyer@redhat.com) +- Add a readiness probe to the Kibana container (skuznets@redhat.com) +- Create logging deployments with non-zero replica counts (skuznets@redhat.com) +- Pulling changes from master branch (ewolinet@redhat.com) +- Adding some missing changes (ewolinet@redhat.com) +- fixing available variables for 2.3.0 (ewolinet@redhat.com) +- Updating pvc generation names (ewolinet@redhat.com) +- updating delete_logging to use modules (ewolinet@redhat.com) +- Pulling in changes from master (ewolinet@redhat.com) +- Decomposing openshift_logging role into subcomponent roles + (ewolinet@redhat.com) +- Fix renaming error with calico template files (djosborne10@gmail.com) + * Tue May 23 2017 Jenkins CD Merge Bot <tdawson@redhat.com> 3.6.80-1 - RPM workaround for the move of cert playbooks (pep@redhat.com) - health check playbooks: relocate and expand (lmeyer@redhat.com) diff --git a/playbooks/byo/openshift-cluster/config.yml b/playbooks/byo/openshift-cluster/config.yml index acf5469bf..fd4a9eb26 100644 --- a/playbooks/byo/openshift-cluster/config.yml +++ b/playbooks/byo/openshift-cluster/config.yml @@ -3,6 +3,19 @@ tags: - always +- name: Verify Requirements + hosts: OSEv3 + roles: + - openshift_health_checker + vars: + - r_openshift_health_checker_playbook_context: "install" + post_tasks: + - action: openshift_health_check + args: + checks: + - disk_availability + - memory_availability + - include: ../../common/openshift-cluster/std_include.yml tags: - always diff --git a/playbooks/common/openshift-checks/health.yml b/playbooks/common/openshift-checks/health.yml index fc0f523d5..1bee460e8 100644 --- a/playbooks/common/openshift-checks/health.yml +++ b/playbooks/common/openshift-checks/health.yml @@ -2,9 +2,10 @@ - name: Run OpenShift health checks hosts: OSEv3 roles: - - openshift_health_checker + - openshift_health_checker + vars: + - r_openshift_health_checker_playbook_context: "health" post_tasks: - - action: openshift_health_check # https://github.com/ansible/ansible/issues/20513 - args: - checks: - - '@health' + - action: openshift_health_check # https://github.com/ansible/ansible/issues/20513 + args: + checks: ['@health'] diff --git a/playbooks/common/openshift-checks/pre-install.yml b/playbooks/common/openshift-checks/pre-install.yml index c8ffc3d91..e01c6f38d 100644 --- a/playbooks/common/openshift-checks/pre-install.yml +++ b/playbooks/common/openshift-checks/pre-install.yml @@ -2,9 +2,10 @@ - hosts: OSEv3 name: run OpenShift pre-install checks roles: - - openshift_health_checker + - openshift_health_checker + vars: + - r_openshift_health_checker_playbook_context: "pre-install" post_tasks: - - action: openshift_health_check # https://github.com/ansible/ansible/issues/20513 - args: - checks: - - '@preflight' + - action: openshift_health_check # https://github.com/ansible/ansible/issues/20513 + args: + checks: ['@preflight'] diff --git a/playbooks/common/openshift-cluster/upgrades/post_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/post_control_plane.yml index 9b76f1dd0..4cf434dab 100644 --- a/playbooks/common/openshift-cluster/upgrades/post_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/post_control_plane.yml @@ -10,7 +10,7 @@ router_image: "{{ openshift.master.registry_url | replace( '${component}', 'haproxy-router' ) | replace ( '${version}', openshift_image_tag ) }}" registry_console_image: "{{ openshift.master.registry_url | replace ( '${component}', 'registry-console') | - replace ( '${version}', openshift.common.short_version ) }}" + replace ( '${version}', 'v' ~ openshift.common.short_version ) }}" pre_tasks: - name: Load lib_openshift modules diff --git a/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml index 0ad934d2d..e10c4c540 100644 --- a/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml @@ -107,8 +107,8 @@ yedit: src: "{{ openshift.common.config_base }}/master/master-config.yaml" key: 'imageConfig.format' - value: "{{ oreg_url }}" - when: oreg_url is defined + value: "{{ oreg_url | default(oreg_url_master) }}" + when: oreg_url is defined or oreg_url_master is defined # Run the upgrade hook prior to restarting services/system if defined: - debug: msg="Running master upgrade hook {{ openshift_master_upgrade_hook }}" diff --git a/playbooks/common/openshift-cluster/validate_hostnames.yml b/playbooks/common/openshift-cluster/validate_hostnames.yml index 48cc03b19..33fc5630f 100644 --- a/playbooks/common/openshift-cluster/validate_hostnames.yml +++ b/playbooks/common/openshift-cluster/validate_hostnames.yml @@ -13,4 +13,6 @@ pause: prompt: "The hostname \"{{ openshift.common.hostname }}\" for \"{{ ansible_nodename }}\" doesn't resolve to an ip address owned by this host. Please set openshift_hostname variable to a hostname that when resolved on the host in question resolves to an IP address matching an interface on this host. This host will fail liveness checks for pods utilizing hostPorts, press ENTER to continue or CTRL-C to abort." seconds: "{{ 10 if openshift_override_hostname_check | default(false) | bool else omit }}" - when: lookupip.stdout not in ansible_all_ipv4_addresses + when: + - lookupip.stdout != '127.0.0.1' + - lookupip.stdout not in ansible_all_ipv4_addresses diff --git a/roles/lib_openshift/library/oc_process.py b/roles/lib_openshift/library/oc_process.py index 330de07eb..15e9c606d 100644 --- a/roles/lib_openshift/library/oc_process.py +++ b/roles/lib_openshift/library/oc_process.py @@ -1450,7 +1450,7 @@ class OCProcess(OpenShiftCLI): if self._template is None: results = self._process(self.name, False, self.params, self.data) if results['returncode'] != 0: - raise OpenShiftCLIError('Error processing template [%s].' % self.name) + raise OpenShiftCLIError('Error processing template [%s]: %s' %(self.name, results)) self._template = results['results']['items'] return self._template diff --git a/roles/lib_openshift/src/class/oc_process.py b/roles/lib_openshift/src/class/oc_process.py index eba9a43cd..62a6bd571 100644 --- a/roles/lib_openshift/src/class/oc_process.py +++ b/roles/lib_openshift/src/class/oc_process.py @@ -30,7 +30,7 @@ class OCProcess(OpenShiftCLI): if self._template is None: results = self._process(self.name, False, self.params, self.data) if results['returncode'] != 0: - raise OpenShiftCLIError('Error processing template [%s].' % self.name) + raise OpenShiftCLIError('Error processing template [%s]: %s' %(self.name, results)) self._template = results['results']['items'] return self._template diff --git a/roles/openshift_health_checker/action_plugins/openshift_health_check.py b/roles/openshift_health_checker/action_plugins/openshift_health_check.py index 03c40b78b..a62e4331e 100644 --- a/roles/openshift_health_checker/action_plugins/openshift_health_check.py +++ b/roles/openshift_health_checker/action_plugins/openshift_health_check.py @@ -25,9 +25,11 @@ class ActionModule(ActionBase): def run(self, tmp=None, task_vars=None): result = super(ActionModule, self).run(tmp, task_vars) + task_vars = task_vars or {} - if task_vars is None: - task_vars = {} + # vars are not supportably available in the callback plugin, + # so record any it will need in the result. + result['playbook_context'] = task_vars.get('r_openshift_health_checker_playbook_context') if "openshift" not in task_vars: result["failed"] = True @@ -46,19 +48,27 @@ class ActionModule(ActionBase): result["checks"] = check_results = {} + user_disabled_checks = [ + check.strip() + for check in task_vars.get("openshift_disable_check", "").split(",") + ] + for check_name in resolved_checks: display.banner("CHECK [{} : {}]".format(check_name, task_vars["ansible_host"])) check = known_checks[check_name] - if check.is_active(task_vars): + if not check.is_active(task_vars): + r = dict(skipped=True, skipped_reason="Not active for this host") + elif check_name in user_disabled_checks: + r = dict(skipped=True, skipped_reason="Disabled by user request") + else: try: r = check.run(tmp, task_vars) except OpenShiftCheckException as e: - r = {} - r["failed"] = True - r["msg"] = str(e) - else: - r = {"skipped": True} + r = dict( + failed=True, + msg=str(e), + ) check_results[check_name] = r diff --git a/roles/openshift_health_checker/callback_plugins/zz_failure_summary.py b/roles/openshift_health_checker/callback_plugins/zz_failure_summary.py index 7bce7f107..64c29a8d9 100644 --- a/roles/openshift_health_checker/callback_plugins/zz_failure_summary.py +++ b/roles/openshift_health_checker/callback_plugins/zz_failure_summary.py @@ -2,6 +2,12 @@ Ansible callback plugin. ''' +# Reason: In several locations below we disable pylint protected-access +# for Ansible objects that do not give us any public way +# to access the full details we need to report check failures. +# Status: disabled permanently or until Ansible object has a public API. +# This does leave the code more likely to be broken by future Ansible changes. + from pprint import pformat from ansible.plugins.callback import CallbackBase @@ -20,38 +26,37 @@ class CallbackModule(CallbackBase): CALLBACK_TYPE = 'aggregate' CALLBACK_NAME = 'failure_summary' CALLBACK_NEEDS_WHITELIST = False + _playbook_file = None def __init__(self): super(CallbackModule, self).__init__() self.__failures = [] + def v2_playbook_on_start(self, playbook): + super(CallbackModule, self).v2_playbook_on_start(playbook) + # re: playbook attrs see top comment # pylint: disable=protected-access + self._playbook_file = playbook._file_name + def v2_runner_on_failed(self, result, ignore_errors=False): super(CallbackModule, self).v2_runner_on_failed(result, ignore_errors) self.__failures.append(dict(result=result, ignore_errors=ignore_errors)) def v2_playbook_on_stats(self, stats): super(CallbackModule, self).v2_playbook_on_stats(stats) - # TODO: update condition to consider a host var or env var to - # enable/disable the summary, so that we can control the output from a - # play. if self.__failures: - self._print_failure_summary() + self._print_failure_details(self.__failures) - def _print_failure_summary(self): - '''Print a summary of failed tasks (including ignored failures).''' + def _print_failure_details(self, failures): + '''Print a summary of failed tasks or checks.''' self._display.display(u'\nFailure summary:\n') - # TODO: group failures by host or by task. If grouped by host, it is - # easy to see all problems of a given host. If grouped by task, it is - # easy to see what hosts needs the same fix. - - width = len(str(len(self.__failures))) + width = len(str(len(failures))) initial_indent_format = u' {{:>{width}}}. '.format(width=width) initial_indent_len = len(initial_indent_format.format(0)) subsequent_indent = u' ' * initial_indent_len subsequent_extra_indent = u' ' * (initial_indent_len + 10) - for i, failure in enumerate(self.__failures, 1): + for i, failure in enumerate(failures, 1): entries = _format_failure(failure) self._display.display(u'\n{}{}'.format(initial_indent_format.format(i), entries[0])) for entry in entries[1:]: @@ -59,11 +64,52 @@ class CallbackModule(CallbackBase): indented = u'{}{}'.format(subsequent_indent, entry) self._display.display(indented) - -# Reason: disable pylint protected-access because we need to access _* -# attributes of a task result to implement this method. -# Status: permanently disabled unless Ansible's API changes. -# pylint: disable=protected-access + failed_checks = set() + playbook_context = None + # re: result attrs see top comment # pylint: disable=protected-access + for failure in failures: + # get context from check task result since callback plugins cannot access task vars + playbook_context = playbook_context or failure['result']._result.get('playbook_context') + failed_checks.update( + name + for name, result in failure['result']._result.get('checks', {}).items() + if result.get('failed') + ) + if failed_checks: + self._print_check_failure_summary(failed_checks, playbook_context) + + def _print_check_failure_summary(self, failed_checks, context): + checks = ','.join(sorted(failed_checks)) + # NOTE: context is not set if all failures occurred prior to checks task + summary = ( + '\n' + 'The execution of "{playbook}"\n' + 'includes checks designed to fail early if the requirements\n' + 'of the playbook are not met. One or more of these checks\n' + 'failed. To disregard these results, you may choose to\n' + 'disable failing checks by setting an Ansible variable:\n\n' + ' openshift_disable_check={checks}\n\n' + 'Failing check names are shown in the failure details above.\n' + 'Some checks may be configurable by variables if your requirements\n' + 'are different from the defaults; consult check documentation.\n' + 'Variables can be set in the inventory or passed on the\n' + 'command line using the -e flag to ansible-playbook.\n' + ).format(playbook=self._playbook_file, checks=checks) + if context in ['pre-install', 'health']: + summary = ( + '\n' + 'You may choose to configure or disable failing checks by\n' + 'setting Ansible variables. To disable those above:\n\n' + ' openshift_disable_check={checks}\n\n' + 'Consult check documentation for configurable variables.\n' + 'Variables can be set in the inventory or passed on the\n' + 'command line using the -e flag to ansible-playbook.\n' + ).format(checks=checks) + # other expected contexts: install, upgrade + self._display.display(summary) + + +# re: result attrs see top comment # pylint: disable=protected-access def _format_failure(failure): '''Return a list of pretty-formatted text entries describing a failure, including relevant information about it. Expect that the list of text entries will be joined @@ -100,11 +146,8 @@ def _format_failed_checks(checks): return stringc(pformat(checks), C.COLOR_ERROR) -# Reason: disable pylint protected-access because we need to access _* -# attributes of obj to implement this function. -# This is inspired by ansible.playbook.base.Base.dump_me. -# Status: permanently disabled unless Ansible's API changes. -# pylint: disable=protected-access +# This is inspired by ansible.playbook.base.Base.dump_me. +# re: play/task/block attrs see top comment # pylint: disable=protected-access def _get_play(obj): '''Given a task or block, recursively tries to find its parent play.''' if hasattr(obj, '_play'): diff --git a/roles/openshift_health_checker/openshift_checks/disk_availability.py b/roles/openshift_health_checker/openshift_checks/disk_availability.py index c2792a0fe..962148cb8 100644 --- a/roles/openshift_health_checker/openshift_checks/disk_availability.py +++ b/roles/openshift_health_checker/openshift_checks/disk_availability.py @@ -27,10 +27,12 @@ class DiskAvailability(NotContainerizedMixin, OpenShiftCheck): def run(self, tmp, task_vars): group_names = get_var(task_vars, "group_names") ansible_mounts = get_var(task_vars, "ansible_mounts") - - min_free_bytes = max(self.recommended_disk_space_bytes.get(name, 0) for name in group_names) free_bytes = self.openshift_available_disk(ansible_mounts) + recommended_min = max(self.recommended_disk_space_bytes.get(name, 0) for name in group_names) + configured_min = int(get_var(task_vars, "openshift_check_min_host_disk_gb", default=0)) * 10**9 + min_free_bytes = configured_min or recommended_min + if free_bytes < min_free_bytes: return { 'failed': True, diff --git a/roles/openshift_health_checker/openshift_checks/memory_availability.py b/roles/openshift_health_checker/openshift_checks/memory_availability.py index 28805dc37..f4e31065f 100644 --- a/roles/openshift_health_checker/openshift_checks/memory_availability.py +++ b/roles/openshift_health_checker/openshift_checks/memory_availability.py @@ -1,6 +1,9 @@ # pylint: disable=missing-docstring from openshift_checks import OpenShiftCheck, get_var +MIB = 2**20 +GIB = 2**30 + class MemoryAvailability(OpenShiftCheck): """Check that recommended memory is available.""" @@ -11,10 +14,12 @@ class MemoryAvailability(OpenShiftCheck): # Values taken from the official installation documentation: # https://docs.openshift.org/latest/install_config/install/prerequisites.html#system-requirements recommended_memory_bytes = { - "masters": 16 * 10**9, - "nodes": 8 * 10**9, - "etcd": 20 * 10**9, + "masters": 16 * GIB, + "nodes": 8 * GIB, + "etcd": 8 * GIB, } + # https://access.redhat.com/solutions/3006511 physical RAM is partly reserved from memtotal + memtotal_adjustment = 1 * GIB @classmethod def is_active(cls, task_vars): @@ -25,19 +30,21 @@ class MemoryAvailability(OpenShiftCheck): def run(self, tmp, task_vars): group_names = get_var(task_vars, "group_names") - total_memory_bytes = get_var(task_vars, "ansible_memtotal_mb") * 10**6 + total_memory_bytes = get_var(task_vars, "ansible_memtotal_mb") * MIB - min_memory_bytes = max(self.recommended_memory_bytes.get(name, 0) for name in group_names) + recommended_min = max(self.recommended_memory_bytes.get(name, 0) for name in group_names) + configured_min = float(get_var(task_vars, "openshift_check_min_host_memory_gb", default=0)) * GIB + min_memory_bytes = configured_min or recommended_min - if total_memory_bytes < min_memory_bytes: + if total_memory_bytes + self.memtotal_adjustment < min_memory_bytes: return { 'failed': True, 'msg': ( - 'Available memory ({available:.1f} GB) ' - 'below recommended value ({recommended:.1f} GB)' + 'Available memory ({available:.1f} GiB) is too far ' + 'below recommended value ({recommended:.1f} GiB)' ).format( - available=float(total_memory_bytes) / 10**9, - recommended=float(min_memory_bytes) / 10**9, + available=float(total_memory_bytes) / GIB, + recommended=float(min_memory_bytes) / GIB, ), } diff --git a/roles/openshift_health_checker/test/action_plugin_test.py b/roles/openshift_health_checker/test/action_plugin_test.py index 2693ae37b..6ebf0ebb2 100644 --- a/roles/openshift_health_checker/test/action_plugin_test.py +++ b/roles/openshift_health_checker/test/action_plugin_test.py @@ -67,6 +67,7 @@ def changed(result): return result.get('changed', False) +# tests whether task is skipped, not individual checks def skipped(result): return result.get('skipped', False) @@ -101,7 +102,20 @@ def test_action_plugin_skip_non_active_checks(plugin, task_vars, monkeypatch): result = plugin.run(tmp=None, task_vars=task_vars) - assert result['checks']['fake_check'] == {'skipped': True} + assert result['checks']['fake_check'] == dict(skipped=True, skipped_reason="Not active for this host") + assert not failed(result) + assert not changed(result) + assert not skipped(result) + + +def test_action_plugin_skip_disabled_checks(plugin, task_vars, monkeypatch): + checks = [fake_check('fake_check', is_active=True)] + monkeypatch.setattr('openshift_checks.OpenShiftCheck.subclasses', classmethod(lambda cls: checks)) + + task_vars['openshift_disable_check'] = 'fake_check' + result = plugin.run(tmp=None, task_vars=task_vars) + + assert result['checks']['fake_check'] == dict(skipped=True, skipped_reason="Disabled by user request") assert not failed(result) assert not changed(result) assert not skipped(result) diff --git a/roles/openshift_health_checker/test/disk_availability_test.py b/roles/openshift_health_checker/test/disk_availability_test.py index 970b474d7..b353fa610 100644 --- a/roles/openshift_health_checker/test/disk_availability_test.py +++ b/roles/openshift_health_checker/test/disk_availability_test.py @@ -42,9 +42,10 @@ def test_cannot_determine_available_disk(ansible_mounts, extra_words): assert word in str(excinfo.value) -@pytest.mark.parametrize('group_names,ansible_mounts', [ +@pytest.mark.parametrize('group_names,configured_min,ansible_mounts', [ ( ['masters'], + 0, [{ 'mount': '/', 'size_available': 40 * 10**9 + 1, @@ -52,6 +53,7 @@ def test_cannot_determine_available_disk(ansible_mounts, extra_words): ), ( ['nodes'], + 0, [{ 'mount': '/', 'size_available': 15 * 10**9 + 1, @@ -59,6 +61,7 @@ def test_cannot_determine_available_disk(ansible_mounts, extra_words): ), ( ['etcd'], + 0, [{ 'mount': '/', 'size_available': 20 * 10**9 + 1, @@ -66,6 +69,15 @@ def test_cannot_determine_available_disk(ansible_mounts, extra_words): ), ( ['etcd'], + 1, # configure lower threshold + [{ + 'mount': '/', + 'size_available': 1 * 10**9 + 1, # way smaller than recommended + }], + ), + ( + ['etcd'], + 0, [{ # not enough space on / ... 'mount': '/', @@ -77,9 +89,10 @@ def test_cannot_determine_available_disk(ansible_mounts, extra_words): }], ), ]) -def test_succeeds_with_recommended_disk_space(group_names, ansible_mounts): +def test_succeeds_with_recommended_disk_space(group_names, configured_min, ansible_mounts): task_vars = dict( group_names=group_names, + openshift_check_min_host_disk_gb=configured_min, ansible_mounts=ansible_mounts, ) @@ -89,9 +102,10 @@ def test_succeeds_with_recommended_disk_space(group_names, ansible_mounts): assert not result.get('failed', False) -@pytest.mark.parametrize('group_names,ansible_mounts,extra_words', [ +@pytest.mark.parametrize('group_names,configured_min,ansible_mounts,extra_words', [ ( ['masters'], + 0, [{ 'mount': '/', 'size_available': 1, @@ -99,7 +113,17 @@ def test_succeeds_with_recommended_disk_space(group_names, ansible_mounts): ['0.0 GB'], ), ( + ['masters'], + 100, # set a higher threshold + [{ + 'mount': '/', + 'size_available': 50 * 10**9, # would normally be enough... + }], + ['100.0 GB'], + ), + ( ['nodes'], + 0, [{ 'mount': '/', 'size_available': 1 * 10**9, @@ -108,6 +132,7 @@ def test_succeeds_with_recommended_disk_space(group_names, ansible_mounts): ), ( ['etcd'], + 0, [{ 'mount': '/', 'size_available': 1, @@ -116,6 +141,7 @@ def test_succeeds_with_recommended_disk_space(group_names, ansible_mounts): ), ( ['nodes', 'masters'], + 0, [{ 'mount': '/', # enough space for a node, not enough for a master @@ -125,6 +151,7 @@ def test_succeeds_with_recommended_disk_space(group_names, ansible_mounts): ), ( ['etcd'], + 0, [{ # enough space on / ... 'mount': '/', @@ -137,9 +164,10 @@ def test_succeeds_with_recommended_disk_space(group_names, ansible_mounts): ['0.0 GB'], ), ]) -def test_fails_with_insufficient_disk_space(group_names, ansible_mounts, extra_words): +def test_fails_with_insufficient_disk_space(group_names, configured_min, ansible_mounts, extra_words): task_vars = dict( group_names=group_names, + openshift_check_min_host_disk_gb=configured_min, ansible_mounts=ansible_mounts, ) diff --git a/roles/openshift_health_checker/test/memory_availability_test.py b/roles/openshift_health_checker/test/memory_availability_test.py index e161a5b9e..4fbaea0a9 100644 --- a/roles/openshift_health_checker/test/memory_availability_test.py +++ b/roles/openshift_health_checker/test/memory_availability_test.py @@ -20,27 +20,42 @@ def test_is_active(group_names, is_active): assert MemoryAvailability.is_active(task_vars=task_vars) == is_active -@pytest.mark.parametrize('group_names,ansible_memtotal_mb', [ +@pytest.mark.parametrize('group_names,configured_min,ansible_memtotal_mb', [ ( ['masters'], + 0, 17200, ), ( ['nodes'], + 0, 8200, ), ( + ['nodes'], + 1, # configure lower threshold + 2000, # too low for recommended but not for configured + ), + ( + ['nodes'], + 2, # configure threshold where adjustment pushes it over + 1900, + ), + ( ['etcd'], - 22200, + 0, + 8200, ), ( ['masters', 'nodes'], + 0, 17000, ), ]) -def test_succeeds_with_recommended_memory(group_names, ansible_memtotal_mb): +def test_succeeds_with_recommended_memory(group_names, configured_min, ansible_memtotal_mb): task_vars = dict( group_names=group_names, + openshift_check_min_host_memory_gb=configured_min, ansible_memtotal_mb=ansible_memtotal_mb, ) @@ -50,39 +65,62 @@ def test_succeeds_with_recommended_memory(group_names, ansible_memtotal_mb): assert not result.get('failed', False) -@pytest.mark.parametrize('group_names,ansible_memtotal_mb,extra_words', [ +@pytest.mark.parametrize('group_names,configured_min,ansible_memtotal_mb,extra_words', [ ( ['masters'], 0, - ['0.0 GB'], + 0, + ['0.0 GiB'], ), ( ['nodes'], + 0, 100, - ['0.1 GB'], + ['0.1 GiB'], + ), + ( + ['nodes'], + 24, # configure higher threshold + 20 * 1024, # enough to meet recommended but not configured + ['20.0 GiB'], + ), + ( + ['nodes'], + 24, # configure higher threshold + 22 * 1024, # not enough for adjustment to push over threshold + ['22.0 GiB'], ), ( ['etcd'], - -1, - ['0.0 GB'], + 0, + 6 * 1024, + ['6.0 GiB'], + ), + ( + ['etcd', 'masters'], + 0, + 9 * 1024, # enough memory for etcd, not enough for a master + ['9.0 GiB'], ), ( ['nodes', 'masters'], + 0, # enough memory for a node, not enough for a master - 11000, - ['11.0 GB'], + 11 * 1024, + ['11.0 GiB'], ), ]) -def test_fails_with_insufficient_memory(group_names, ansible_memtotal_mb, extra_words): +def test_fails_with_insufficient_memory(group_names, configured_min, ansible_memtotal_mb, extra_words): task_vars = dict( group_names=group_names, + openshift_check_min_host_memory_gb=configured_min, ansible_memtotal_mb=ansible_memtotal_mb, ) check = MemoryAvailability(execute_module=fake_execute_module) result = check.run(tmp=None, task_vars=task_vars) - assert result['failed'] + assert result.get('failed', False) for word in 'below recommended'.split() + extra_words: assert word in result['msg'] diff --git a/roles/openshift_hosted_templates/files/v3.6/enterprise/registry-console.yaml b/roles/openshift_hosted_templates/files/v3.6/enterprise/registry-console.yaml index 8fe02444e..8bf98ba41 100644 --- a/roles/openshift_hosted_templates/files/v3.6/enterprise/registry-console.yaml +++ b/roles/openshift_hosted_templates/files/v3.6/enterprise/registry-console.yaml @@ -103,9 +103,9 @@ parameters: - description: 'Specify "registry/repository" prefix for container image; e.g. for "registry.access.redhat.com/openshift3/registry-console:latest", set prefix "registry.access.redhat.com/openshift3/"' name: IMAGE_PREFIX value: "registry.access.redhat.com/openshift3/" - - description: 'Specify image version; e.g. for "registry.access.redhat.com/openshift3/registry-console:3.6", set version "3.6"' + - description: 'Specify image version; e.g. for "registry.access.redhat.com/openshift3/registry-console:v3.6", set version "v3.6"' name: IMAGE_VERSION - value: "3.6" + value: "v3.6" - description: "The public URL for the Openshift OAuth Provider, e.g. https://openshift.example.com:8443" name: OPENSHIFT_OAUTH_PROVIDER_URL required: true diff --git a/roles/openshift_logging/filter_plugins/openshift_logging.py b/roles/openshift_logging/filter_plugins/openshift_logging.py index 44b0b2d48..eac086e81 100644 --- a/roles/openshift_logging/filter_plugins/openshift_logging.py +++ b/roles/openshift_logging/filter_plugins/openshift_logging.py @@ -42,7 +42,7 @@ def map_from_pairs(source, delim="="): if source == '': return dict() - return dict(source.split(delim) for item in source.split(",")) + return dict(item.split(delim) for item in source.split(",")) # pylint: disable=too-few-public-methods diff --git a/roles/openshift_logging_elasticsearch/defaults/main.yml b/roles/openshift_logging_elasticsearch/defaults/main.yml index 7923059da..c0b5d394e 100644 --- a/roles/openshift_logging_elasticsearch/defaults/main.yml +++ b/roles/openshift_logging_elasticsearch/defaults/main.yml @@ -5,10 +5,10 @@ openshift_logging_elasticsearch_image_version: "{{ openshift_hosted_logging_depl openshift_logging_elasticsearch_image_pull_secret: "{{ openshift_hosted_logging_image_pull_secret | default('') }}" openshift_logging_elasticsearch_namespace: logging -openshift_logging_elasticsearch_nodeselector: "" -openshift_logging_elasticsearch_cpu_limit: 100m -openshift_logging_elasticsearch_memory_limit: 512Mi -openshift_logging_elasticsearch_recover_after_time: 5m +openshift_logging_elasticsearch_nodeselector: "{{ openshift_logging_es_nodeselector | default('') }}" +openshift_logging_elasticsearch_cpu_limit: 1000m +openshift_logging_elasticsearch_memory_limit: "{{ openshift_logging_es_memory_limit | default('1Gi') }}" +openshift_logging_elasticsearch_recover_after_time: "{{ openshift_logging_es_recover_after_time | default('5m') }}" openshift_logging_elasticsearch_replica_count: 1 diff --git a/roles/openshift_master/README.md b/roles/openshift_master/README.md index c3300a7ef..e5362105c 100644 --- a/roles/openshift_master/README.md +++ b/roles/openshift_master/README.md @@ -15,17 +15,18 @@ Role Variables From this role: -| Name | Default value | | -|-------------------------------------|-----------------------|--------------------------------------------------| -| openshift_master_debug_level | openshift_debug_level | Verbosity of the debug logs for master | +| Name | Default value | | +|-------------------------------------|-----------------------|-------------------------------------------------------------------------------| +| openshift_master_debug_level | openshift_debug_level | Verbosity of the debug logs for master | | openshift_node_ips | [] | List of the openshift node ip addresses to pre-register when master starts up | -| oreg_url | UNDEF | Default docker registry to use | -| openshift_master_api_port | UNDEF | | -| openshift_master_console_port | UNDEF | | -| openshift_master_api_url | UNDEF | | -| openshift_master_console_url | UNDEF | | -| openshift_master_public_api_url | UNDEF | | -| openshift_master_public_console_url | UNDEF | | +| oreg_url | UNDEF | Default docker registry to use | +| oreg_url_master | UNDEF | Default docker registry to use, specifically on the master | +| openshift_master_api_port | UNDEF | | +| openshift_master_console_port | UNDEF | | +| openshift_master_api_url | UNDEF | | +| openshift_master_console_url | UNDEF | | +| openshift_master_public_api_url | UNDEF | | +| openshift_master_public_console_url | UNDEF | | From openshift_common: diff --git a/roles/openshift_master_facts/tasks/main.yml b/roles/openshift_master_facts/tasks/main.yml index 79f054b42..ef8dcd5fd 100644 --- a/roles/openshift_master_facts/tasks/main.yml +++ b/roles/openshift_master_facts/tasks/main.yml @@ -74,7 +74,7 @@ ldap_ca: "{{ openshift_master_ldap_ca | default(lookup('file', openshift_master_ldap_ca_file) if openshift_master_ldap_ca_file is defined else None) }}" openid_ca: "{{ openshift_master_openid_ca | default(lookup('file', openshift_master_openid_ca_file) if openshift_master_openid_ca_file is defined else None) }}" request_header_ca: "{{ openshift_master_request_header_ca | default(lookup('file', openshift_master_request_header_ca_file) if openshift_master_request_header_ca_file is defined else None) }}" - registry_url: "{{ oreg_url | default(None) }}" + registry_url: "{{ oreg_url_master | default(oreg_url) | default(None) }}" oauth_grant_method: "{{ openshift_master_oauth_grant_method | default(None) }}" sdn_cluster_network_cidr: "{{ osm_cluster_network_cidr | default(None) }}" sdn_host_subnet_length: "{{ osm_host_subnet_length | default(None) }}" diff --git a/roles/openshift_metrics/tasks/main.yaml b/roles/openshift_metrics/tasks/main.yaml index 9af10a849..5d8506a73 100644 --- a/roles/openshift_metrics/tasks/main.yaml +++ b/roles/openshift_metrics/tasks/main.yaml @@ -1,5 +1,5 @@ --- -- local_action: shell rpm -q python-passlib || echo not installed +- local_action: shell python -c 'import passlib' 2>/dev/null || echo not installed register: passlib_result - name: Check that python-passlib is available on the control host diff --git a/roles/openshift_node/README.md b/roles/openshift_node/README.md index b69b60c1d..fb0b494da 100644 --- a/roles/openshift_node/README.md +++ b/roles/openshift_node/README.md @@ -15,10 +15,11 @@ Role Variables -------------- From this role: -| Name | Default value | | -|------------------------------------------|-----------------------|--------------------------------------------------------| -| openshift_node_debug_level | openshift_debug_level | Verbosity of the debug logs for node | -| oreg_url | UNDEF (Optional) | Default docker registry to use | +| Name | Default value | | +|----------------------------|-----------------------|----------------------------------------------------------| +| openshift_node_debug_level | openshift_debug_level | Verbosity of the debug logs for node | +| oreg_url | UNDEF (Optional) | Default docker registry to use | +| oreg_url_node | UNDEF (Optional) | Default docker registry to use, specifically on the node | From openshift_common: diff --git a/roles/openshift_node/tasks/main.yml b/roles/openshift_node/tasks/main.yml index 54ba5ba01..a8beaa060 100644 --- a/roles/openshift_node/tasks/main.yml +++ b/roles/openshift_node/tasks/main.yml @@ -22,7 +22,7 @@ iptables_sync_period: "{{ openshift_node_iptables_sync_period | default(None) }}" kubelet_args: "{{ openshift_node_kubelet_args | default(None) }}" labels: "{{ lookup('oo_option', 'openshift_node_labels') | default( openshift_node_labels | default(none), true) }}" - registry_url: "{{ oreg_url | default(none) }}" + registry_url: "{{ oreg_url_node | default(oreg_url) | default(None) }}" schedulable: "{{ openshift_schedulable | default(openshift_scheduleable) | default(None) }}" sdn_mtu: "{{ openshift_node_sdn_mtu | default(None) }}" storage_plugin_deps: "{{ osn_storage_plugin_deps | default(None) }}" diff --git a/roles/openshift_node_upgrade/tasks/main.yml b/roles/openshift_node_upgrade/tasks/main.yml index 7231bdb9d..d44839d69 100644 --- a/roles/openshift_node_upgrade/tasks/main.yml +++ b/roles/openshift_node_upgrade/tasks/main.yml @@ -92,8 +92,8 @@ yedit: src: "{{ openshift.common.config_base }}/node/node-config.yaml" key: 'imageConfig.format' - value: "{{ oreg_url }}" - when: oreg_url is defined + value: "{{ oreg_url | default(oreg_url_node) }}" + when: oreg_url is defined or oreg_url_node is defined # https://docs.openshift.com/container-platform/3.4/admin_guide/overcommit.html#disabling-swap-memory - name: Check for swap usage diff --git a/roles/openshift_storage_glusterfs/filter_plugins/openshift_storage_glusterfs.py b/roles/openshift_storage_glusterfs/filter_plugins/openshift_storage_glusterfs.py index 88801e487..a86c96df7 100644 --- a/roles/openshift_storage_glusterfs/filter_plugins/openshift_storage_glusterfs.py +++ b/roles/openshift_storage_glusterfs/filter_plugins/openshift_storage_glusterfs.py @@ -8,7 +8,7 @@ def map_from_pairs(source, delim="="): if source == '': return dict() - return dict(source.split(delim) for item in source.split(",")) + return dict(item.split(delim) for item in source.split(",")) # pylint: disable=too-few-public-methods diff --git a/roles/openshift_version/tasks/main.yml b/roles/openshift_version/tasks/main.yml index f2f4d16f0..b35a3fa3c 100644 --- a/roles/openshift_version/tasks/main.yml +++ b/roles/openshift_version/tasks/main.yml @@ -117,7 +117,7 @@ debug: msg: > openshift_image_tag is used for containerized installs. If you are trying to - specify an image for a non-container install see oreg_url. + specify an image for a non-container install see oreg_url or oreg_url_master or oreg_url_node. when: - not is_containerized | bool - openshift_image_tag is defined |