summaryrefslogtreecommitdiffstats
path: root/roles/openshift_logging/tasks
Commit message (Collapse)AuthorAgeFilesLines
...
| * | Add logging es prometheus endpointJeff Cantrill2017-10-032-2/+3
| |/
* | Merge pull request #5543 from ewolinetz/bz1495150OpenShift Merge Robot2017-10-041-4/+0
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Automatic merge from submit-queue. Removing setting pvc size and dynamic to remove looped var setting If we don't set openshift_logging_es_pvc_size but have `openshift_logging_es_pvc_dynamic=True` we see the variable openshift_logging_elasticsearch_pvc_size is set recursively as itself. Addresses: https://bugzilla.redhat.com/show_bug.cgi?id=1495150 https://bugzilla.redhat.com/show_bug.cgi?id=1496202
| * | Removing setting pvc size and dynamic to remove looped var settingEric Wolinetz2017-09-261-4/+0
| | |
* | | Reverting using uninstall variables for logging and metricsEric Wolinetz2017-09-291-1/+1
| |/ |/|
* | logging: introducing event routerJan Wozniak2017-09-262-0/+14
|/ | | | | | | | - eventrouter has its own role 'openshift_logging_eventrouter' written as a template - 'openshift_logging_install_eventrouter' controls whether it gets included in the playbook or not - deployed by default to 'default' namespace
* Updating default behavior for installing metrics and logging. Separating out ↵Eric Wolinetz2017-09-221-4/+5
| | | | uninstall to own variable
* Bug 1491636 - honor openshift_logging_es_ops_nodeselectorJan Wozniak2017-09-141-0/+2
|
* Adding deprecation checks to ansible_syntaxRussell Teague2017-09-011-1/+1
|
* Merge pull request #5189 from wozniakjan/logging_purge_uninstallOpenShift Bot2017-08-311-0/+10
|\ | | | | Merged by openshift-bot
| * Bug 1467265 - logging: add 'purge' option with uninstallJan Wozniak2017-08-241-0/+10
| | | | | | | | | | | | | | | | | | | | | | The common uninstall would keep PVC around to prevent unwanted data loss during reinstalls. There is a scenario where the user may want complete removal of EFK stack including PVC. This change adds an optional `openshift_logging_purge_logging` variable to completely remove the EFK stack. By default, this is set to 'False' and only works if user sets `openshift_logging_install_logging` to 'False' to initiate the uninstallation process.
* | Merge pull request #5219 from jcantrill/bz1480988_kibana_ops_linksScott Dodson2017-08-292-0/+18
|\ \ | | | | | | annotate the infra projects for logging to fix bz1480988
| * | annotate the infra projects for logging to fix bz1480988Jeff Cantrill2017-08-282-0/+18
| | |
* | | Merge pull request #5044 from wozniakjan/logging_role_based_image_versionsOpenShift Bot2017-08-251-12/+0
|\ \ \ | |/ / |/| | Merged by openshift-bot
| * | Bug 1471322: logging roles based image versionsJan Wozniak2017-08-181-12/+0
| | | | | | | | | | | | | | | | | | Allowing to specify an image version for each logging component https://bugzilla.redhat.com/show_bug.cgi?id=1471322
* | | Passing memory and cpu limit for ops ES installEric Wolinetz2017-08-221-0/+4
| |/ |/|
* | Remove additional 'restart master' handler references.Andrew Butcher2017-08-091-1/+0
|/
* Updating how storage type is determined, adding bool filter in ↵ewolinetz2017-07-241-4/+12
| | | | openshift_logging_elasticsearch
* Bug 1466152 - Json-file log driver: Neither ↵Rich Megginson2017-07-061-1/+0
| | | | | | | | | | "openshift_logging_fluentd_use_journal=false" nor omitted collects the log entries https://bugzilla.redhat.com/show_bug.cgi?id=1466152 Do not set openshift_logging_fluentd_use_journal or USE_JOURNAL at all unless it is explicitly set as an ansible param. It is almost always better to let fluentd figure out which log driver docker is using.
* fix 1466680. Fix logging deploying to the specified namespaceJeff Cantrill2017-06-301-0/+4
|
* Fixing ops storage options being passed to openshift_logging_elasticsearch ↵ewolinetz2017-06-291-10/+12
| | | | role fixing default ops pv selector
* fix es routes for new logging rolesRich Megginson2017-06-071-0/+12
| | | | | | | | | | | | port the code that creates the external Elasticsearch routes to the new logging roles Have to suppress this error message: SSL Problem illegal change cipher spec msg, conn state = 6, handshake state = 1 which is coming from the router health check, until https://github.com/openshift/origin/issues/14515 is fixed - otherwise, the es log is spammed relentlessly
* fix up openshift-ansible for use with 'oc cluster up'Jeff Cantrill2017-06-062-2/+2
|
* Merge pull request #4327 from ewolinetz/fix_logging_secretOpenShift Bot2017-06-062-4/+6
|\ | | | | Merged by openshift-bot
| * Updating Kibana-proxy secret key name, fixing deleting secrets, fixed extra ↵ewolinetz2017-05-312-4/+6
| | | | | | | | ES dc creation
* | Merge pull request #4232 from richm/mux-namespaces-docsOpenShift Bot2017-06-021-1/+1
|\ \ | | | | | | Merged by openshift-bot
| * | add mux docs; allow to specify mux namespacesRich Megginson2017-05-311-1/+1
| |/ | | | | | | | | | | | | | | | | | | | | This adds the necessary documentation for the mux parameters and behavior. This also adds new parameters which allow to specify the namespaces mux must create by default, and which can be optionally added. This casts openshift_logging_use_mux to bool wherever it is used as a boolean. Also - use oc_service `labels` and `external_ips` which are now available.
* | fix curator host, port params; remove curator es volumesRich Megginson2017-05-311-0/+4
| |
* | Revert "Merge pull request #4271 from DG-i/master"Steve Kuznetsov2017-05-312-2/+0
|/ | | | | | | This reverts commit 6b9e05d16a935d632b83eda601538bff2353e4fb, reversing changes made to 240b923c95e04e9467c607c502c857dd86d14179. Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
* Adding default value for openshift_hosted_logging_storage_kindewolinetz2017-05-301-4/+4
|
* Merge pull request #4280 from ewolinetz/logging_pvc_fixOpenShift Bot2017-05-301-4/+7
|\ | | | | Merged by openshift-bot
| * Updating to generate PVC when storage type is passed in as nfsewolinetz2017-05-251-4/+7
| |
* | disable become for local actionsMathias Merscher2017-05-242-0/+2
|/
* Fixing tux warnings and some final clean upewolinetz2017-05-231-7/+9
|
* fixing available variables for 2.3.0ewolinetz2017-05-221-0/+8
|
* Updating pvc generation namesewolinetz2017-05-221-3/+20
|
* updating delete_logging to use modulesewolinetz2017-05-221-62/+47
|
* Pulling in changes from masterewolinetz2017-05-2225-1663/+46
|
* Decomposing openshift_logging role into subcomponent rolesewolinetz2017-05-224-119/+161
|
* Fix yamllint problemsScott Dodson2017-05-191-8/+8
| | | | I should've noticed before merging...
* add ability to expose Elasticsearch as an external routeRich Megginson2017-05-183-0/+146
| | | | | | | | | | | | | | This adds the ability to expose Elastisearch as a route outside of the cluster. - `openshift_logging_es_allow_external`: True (default is False) - if this is True, Elasticsearch will be exposed as a Route - `openshift_logging_es_ops_hostname`: The external facing hostname to use for the route and the TLS server certificate (default is "es." + `openshift_master_default_subdomain`) There are other similar parameters for the TLS server cert, key, and CA cert. There are other similar parameters for when the OPS cluster is deployed e.g. `openshift_logging_es_ops_allow_external`, etc.
* Don't escalate privileges in local tmpdir creationSteve Kuznetsov2017-05-161-0/+1
| | | | | | | | | | When running `ansible-playbook` as a non-root user, if a local temporary directory is created without explicitly not escalating privileges, the directory will not be readable to the user running the playbook on the controller host, which will lead to tasks like `fetch` failing to place files into the directory. Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
* Merge pull request #4026 from ewolinetz/cherry_picks_15OpenShift Bot2017-05-084-11/+28
|\ | | | | Merged by openshift-bot
| * Updating logging and metrics to restart api, ha and controllers when ↵ewolinetz2017-04-282-4/+7
| | | | | | | | updating master config
| * Adding defaults for es_indicesewolinetz2017-04-271-2/+2
| |
| * Updating logic for generating pvcs and their counts to prevent reuse when ↵ewolinetz2017-04-272-7/+21
| | | | | | | | looping
* | mux startup is broken without this fixRich Megginson2017-05-042-2/+2
| |
* | Correctly setting the primary and replica shard count settingsewolinetz2017-05-033-10/+2
| |
* | Making mux with_items list evaluate as empty if didnt get objects beforeewolinetz2017-05-022-2/+2
| |
* | Merge pull request #4028 from ewolinetz/logging_project_ns_labelsOpenShift Bot2017-05-021-11/+30
|\ \ | | | | | | Merged by openshift-bot
| * | Adding way to add labels and nodeselectors to logging projectewolinetz2017-04-271-11/+30
| |/