summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* * repos role renamed to openshift_reposJhon Honce2015-03-241-1/+1
|
* Revert "Jwhonce wip/cluster"Jhon Honce2015-03-247-184/+9
|
* gce inventory/playbook updates for node registration changesJason DeTiberus2015-03-246-3/+19
|
* remove openshift_hostname_workaround var for openshift_common, rather rely ↵Jason DeTiberus2015-03-245-9/+0
| | | | on inventory/playbook variables for openshift_hostname
* node registration changesJason DeTiberus2015-03-2413-122/+282
| | | | | | | | | | | | | | - Remove default value for openshift_hostname and make it required - Remove workarounds that are no longer needed - Remove resources parameter from openshift_register_node module - pre-create node certificates for each node before registering node - distribute created node certificates to each node - Move node registration logic to a new openshift_register_nodes role - This is because we now have to run the steps on a master as opposed to on the nodes like we were previously doing. - Rename openshift_register_node module to kubernetes_register_node, one more step to genericizing enough for upstreaming, however there are still plenty of openshift specific commands that still need to be genericized.
* Rename repos role to openshift_reposJason DeTiberus2015-03-241-1/+1
| | | | | | | | | - Rename repos role to openshift_repos - Make openshift_repos a dependency of openshift_common - Add README and metadata for openshift_repos - Playbook updates for role rename - Verify libselinux-python is installed, otherwise some of the bulit-in modules we use fail
* * Updates from code reviewsJhon Honce2015-03-244-4/+14
|
* remove os_firewall creation of DOCKER chainJason DeTiberus2015-03-241-8/+0
|
* Use docker as package name instead of docker-ioJason DeTiberus2015-03-241-1/+1
|
* Add new role os_env_extras_node that is a subset of the docker roleJason DeTiberus2015-03-245-1/+183
| | | | | | | - Does not install or start docker, since the openshift-node role will handle that for us - Only add root to the dockerroot group and configures the enter-container script.
* * Replace asserts with raisesJhon Honce2015-03-241-2/+2
|
* * Updates from code reviewsJhon Honce2015-03-241-5/+8
|
* Add verify_chain action to os_firewall_manage_iptables moduleJason DeTiberus2015-03-242-35/+47
| | | | | | - Add verify_chain action to os_firewall_manage_iptables module - Update os_firewall module to use os_firewall_manage_iptables for creating the DOCKER chain.
* * Add DOCKER chain to iptablesJhon Honce2015-03-243-6/+15
|
* use more specific variable names in gce/openshift-cluster/launch.ymlJason DeTiberus2015-03-241-6/+6
|
* replace oo_hosts_to_config with oo_nodes_to_config and oo_masters_to_configJason DeTiberus2015-03-247-17/+16
|
* Fix openshift_master_ips and openshift_master_public_ips resolutionJason DeTiberus2015-03-241-2/+1
| | | | | | | | | | | - don't use set_fact on localhost for openshift_master_ips and openshift_master_public_ips - we are only using it for the configure play - move definition to vars section of configure play - otherwise we'd have to set openshift_master_ips and openshift_master_public_ips from hostvars['localhost'] and since we aren't refrerencing it anywhere else, might as well just do it in vars instead of set_fact on locahost.
* add repos role to gce cluster launch so that we are applying ↵Jason DeTiberus2015-03-242-3/+1
| | | | os_update_latest after repo config
* Use env for gce paramsJason DeTiberus2015-03-241-6/+6
|
* Use ansible playbook to initialize openshift clusterJhon Honce2015-03-242-0/+4
| | | | | * Added playbooks/gce/openshift-cluster * Added bin/cluster (will replace cluster.sh)
* Various fixesJason DeTiberus2015-03-245-89/+133
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - playbooks/gce/openshift-cluster: - Remove some stray debugging statements - Some minor formatting fixes - removing un-necessary quotes - cleaning up some jinja templates for readability - add a play to the launch playbook to apply the os_update_latest role on all hosts in the new environment - improve setting groups and gce_public_ip when using add_host module - set gce_public_ip as a variable for the host using the returned gce instance_data - add a group for each tag configured on the host (pre-pending tag_ to the tag name) - update the openshift-master/config.yml and openshift-node/config.yml includes to use the tag_env-host-type groups - openshift-{master,node}/config.yml - Some cleanup - remove some extraneous quotes - remove connection: ssh from remote hosts, since it is the default - remove user: root and instead set ansible_ssh_user in inventory/gce/group_vars/all - set openshift_public_ip and openshift_env to templated values in inventory/gce/group_vars/all as well - no longer set openshift_node_ips for the master host, since nodes will register themselves now when they are configured (prevent reboot on adding nodes) - move setting openshift_master_ips and openshift_public_master_ips using set_fact and instead use the vars: of the 'Configure Instances' play
* add roles symlink to playbooks/gce/openshift-cluster to allow launch to call ↵Jason DeTiberus2015-03-241-0/+1
| | | | os_update_latest role
* Move yum update * to new os_update_latest roleJason DeTiberus2015-03-242-3/+3
|
* add oo_prepend_strings_in_list filterJason DeTiberus2015-03-241-44/+62
|
* Add spacing to implicit string concatenation for python backwards compatibilityJason DeTiberus2015-03-241-4/+4
|
* Use ansible playbook to initialize openshift clusterJhon Honce2015-03-2412-7/+253
| | | | | * Added playbooks/gce/openshift-cluster * Added bin/cluster (will replace cluster.sh)
* Merge pull request #118 from liangxia/masterThomas Wiest2015-03-242-2/+2
|\ | | | | minor fix
| * minor fixliangxia2015-03-192-2/+2
| |
* | Merge pull request #117 from detiber/repoRenameThomas Wiest2015-03-2419-6/+61
|\ \ | | | | | | Rename repos role to openshift_repos
| * | Rename repos role to openshift_reposJason DeTiberus2015-03-1819-6/+61
| | | | | | | | | | | | | | | | | | | | | | | | | | | - Rename repos role to openshift_repos - Make openshift_repos a dependency of openshift_common - Add README and metadata for openshift_repos - Playbook updates for role rename - Verify libselinux-python is installed, otherwise some of the bulit-in modules we use fail
* | | Update README_GCE.mdDan McPherson2015-03-231-1/+1
| | |
* | | Update README_GCE.mdDan McPherson2015-03-231-1/+1
| |/ |/|
* | Merge pull request #116 from detiber/bugSquashingThomas Wiest2015-03-186-149/+336
|\ \ | |/ |/| Bug squashing
| * Register node fixesJason DeTiberus2015-03-183-142/+333
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Set --hostname flag in node config in openshift_node role - Support some additional node attributes in openshift_node role - podCIDR - labels - annotations - Support both output types for openshift ex config view in openshift_register_node module - Support multiple api versions in openshift_register_node module - Support additional attributes in openshift_register_node module - annotations - labels - pod_cidr - external_ips (v1beta3, will be available after next kube rebase) - internal_ips (v1beta3, will be available after next kube rebase) - hostnames (v1beta3, will be available after next kube rebase) - external_id (v1beta3, will be available after next kube rebase)
| * Fix hostname handlingJason DeTiberus2015-03-183-7/+3
|/ | | | | - always set hostname if hostname does not match openshift_hostname - Use local IP instead of public IP as hostname for workaround
* Merge pull request #114 from twiest/prThomas Wiest2015-03-134-14/+18
|\ | | | | Renamed AnsibleUtil to AwsUtil. Fixed bug in AwsUtil for hosts without environment set.
| * made the bug fix more python-esqThomas Wiest2015-03-131-1/+1
| |
| * fixed bug in AwsUtil that was causing it to blow up on hosts without an ↵Thomas Wiest2015-03-131-0/+4
| | | | | | | | environment set.
| * renamed AnsibleUtil to AwsUtil because that's what it really is.Thomas Wiest2015-03-134-14/+14
|/
* Merge pull request #112 from detiber/masterWorkaroundThomas Wiest2015-03-131-0/+17
|\ | | | | Add workaround for openshift-master startup timeout
| * Add workaround for openshift-master startup timeoutJason DeTiberus2015-03-131-0/+17
| |
* | Merge pull request #113 from detiber/kubeFormatThomas Wiest2015-03-131-5/+11
|\ \ | |/ |/| Update openshift_register_node module for changes in .kubeconfig format ...
| * Update openshift_register_node module for changes in .kubeconfig format ↵Jason DeTiberus2015-03-131-5/+11
|/ | | | following latest kubernetes rebase
* Merge pull request #111 from twiest/opsshThomas Wiest2015-03-112-1/+179
|\ | | | | added opssh
| * added opsshThomas Wiest2015-03-112-1/+179
|/
* Merge pull request #70 from lhuard1A/fix_awsThomas Wiest2015-03-103-5/+11
|\ | | | | Fix AWS playbooks
| * Fix AWS playbooksLénaïc Huard2015-03-103-5/+11
| |
* | Merge pull request #110 from detiber/nodeResourcesThomas Wiest2015-03-093-2/+6
|\ \ | | | | | | Allow for overriding node resources
| * | Allow for overriding node resourcesJason DeTiberus2015-03-093-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - add variable openshift_node_resources to openshift_node role - set default value for openshift_node_resources to { capacity: { cpu: ,memory: }} - If cpu is not set, then the default value will be chosen by the openshift_register_node module (num logical cpus) - If memory is not set, then the default value will be chosen by the openshift_register_node module (75% MemTotal according to /proc/meminfo)
* | | Merge pull request #109 from twiest/prThomas Wiest2015-03-094-5/+5
|\ \ \ | |/ / |/| | fixed github repo name in READMEs