Changelog for velum-4.0.0+dev+git_r970_e7df5eb43377899ccfa1072fc1542924d4cebf6d-211.21.x86_64.rpm :

* Tue Feb 19 2019 containers-bugownerAATTsuse.de- Commit 5a239fd by Florian Bergmann fbergmannAATTsuse.de Do not allow \'.\' or \'/\' symbols in system certificate names. This way when salt extracts the name to construct the certificate path, it will not be able to break out of the folder for trusted certificates. This is a first layer fix for bnc#1121447.
* Mon Feb 11 2019 containers-bugownerAATTsuse.de- Commit a6e9eaf by Miquel Sabaté Solà msabateAATTsuse.com oidc: fixed instructions for manual configuration bsc#1123711 Signed-off-by: Miquel Sabaté Solà
* Wed Jan 09 2019 containers-bugownerAATTsuse.de- Commit bcec0b3 by dmaiocchi dmaiocchiAATTsuse.com remove hack, not anymore needed
* Tue Dec 04 2018 containers-bugownerAATTsuse.de- Commit 71c3c23 by James Mason jmasonAATTsuse.com [BSC#1101219] Use MSI authorization in Azure Supplying credentials is no longer required for starting instances in Azure; Azure MSI acts like machine accounts in AWS IAM, for example, where the instance is granted permission to interact with the framework (for starting cluster instances, in our case).
* Tue Dec 04 2018 containers-bugownerAATTsuse.de- Commit a1ac59f by Maximilian Meister mmeisterAATTsuse.de enable update channels bsc#1109785 Signed-off-by: Maximilian Meister
* Tue Dec 04 2018 containers-bugownerAATTsuse.de- Commit c741705 by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1117152: Do not parse the return code but the JSON response.
* Thu Nov 22 2018 containers-bugownerAATTsuse.de- Commit af5f4a4 by Vítor Avelino vavelinoAATTsuse.com ldap: minor fix field value example Signed-off-by: Vítor Avelino bsc#1109293
* Thu Nov 15 2018 containers-bugownerAATTsuse.de- Commit 1e69eb1 by Vítor Avelino vavelinoAATTsuse.com js: added eslint ESLint rules based on Portus have been added and most of the issues were fixed. The dashboard.js file was the only one postponed because there are pending PRs touching it and since there are a lot of changes, it would cause a lot of conflicts. Also moved ESLint, rubocop and brakeman to run before specs to fail faster if there\'s a styling/security issue. Signed-off-by: Vítor Avelino
* Mon Nov 12 2018 containers-bugownerAATTsuse.de- Commit 1e69eb1 by Vítor Avelino vavelinoAATTsuse.com js: added eslint ESLint rules based on Portus have been added and most of the issues were fixed. The dashboard.js file was the only one postponed because there are pending PRs touching it and since there are a lot of changes, it would cause a lot of conflicts. Also moved ESLint, rubocop and brakeman to run before specs to fail faster if there\'s a styling/security issue. Signed-off-by: Vítor Avelino
* Wed Nov 07 2018 containers-bugownerAATTsuse.de- Commit e0ab204 by Danny Sauer dsauerAATTsuse.com Complete OIDC connector support
* Fixed oidc database migration
* Suppress autocomplete
* Remove unecessary exception trap
* Added small-format help text
* Pull callback URL from pillar
* Move callback URL to a hidden field (User should not be interacting with callback URL, but may need to use it during client setup.)
* Fixed update and delete methods
* Add http url validator and use on oidc connector
* Rename Endpoint URL label to Provider URL
* Working OIDC Provider validation
* Rename \"oidc server\" to \"oidc provider\"
* Clean up code for Rubocop compliance
* Moved formerly stand-alone documentation and test structure into source tree
* Hide the dex.db
* Prevent local config from being comitted
* Add notices which still don\'t display :/
* Bump schema version so oidc table gets built
* Implement test cases
* Remove extraneous transaction wrappers
* Add validation skip trait to oidc connector factory
* Fix pillar OIDC tests which always passed
* better salt pillar test
* Complete rewrite of example dex config
* Add pillar test w/ basic auth
* Modify former \"basic auth\" form to be more descriptive
* Suppress client secret from logging
* suppress ssl cert checking when adding connector (will break orchestration for self-signed cert)
* Wed Oct 31 2018 containers-bugownerAATTsuse.de- Commit 820cfb8 by Maximilian Meister mmeisterAATTsuse.de followup fix for bsc#1108740 we need to check the admin for `tx_update_reboot_needed`, not the minions Signed-off-by: Maximilian Meister
* Wed Oct 31 2018 containers-bugownerAATTsuse.de- Commit b09dc0a by Michal Jura mjuraAATTsuse.com Fix alter_time format when deleting old data (bsc#1112967)
* Mon Oct 29 2018 containers-bugownerAATTsuse.de- Commit 778d447 by Joachim Gleissner jgleissnerAATTsuse.com cloud_cluster: fix current_cluster_size (bsc#1113316)
* Tue Oct 23 2018 containers-bugownerAATTsuse.de- Commit 302f59e by Chang Lee cleeAATTsuse.com Velum: Add more log filters for user\'s secrets, pass, and tokens bug - 1111848 bsc#1111848
* Fri Oct 19 2018 containers-bugownerAATTsuse.de- Commit aa9e2a5 by James Mason jmasonAATTsuse.com [bsc#1109720] test for salt \'alter_time\' attributes Incorporate the `alter_time` patch into the schema by default, and test that conditions are met. Instead of having to remember to update a non-obvious patch when adding a migration, developers must remember not to include changes to the affected schema lines, and will be assisted by a spec that verifies the proper conditions are maintained. BONUS: test for the conditions _without_ analyzing the text content of the schema file, so that whenever velum moves to Rails 5, we can incorporate the proper upstream implementation[1] [1] https://stackoverflow.com/a/40647191
* Thu Oct 11 2018 containers-bugownerAATTsuse.de- Commit a17a674 by Michal Jura mjuraAATTsuse.com [CPI] OpenStack Cloud Provider config file doesn\'t exist (bsc#1111361)
* Wed Oct 10 2018 containers-bugownerAATTsuse.de- Commit a271085 by James Mason jmasonAATTsuse.com Remove code climate badges from README
* Tue Oct 09 2018 containers-bugownerAATTsuse.de- Commit 422f207 by Kiall Mac Innes kiallAATTmacinnes.ie Remove Code Climate config
* Mon Oct 08 2018 containers-bugownerAATTsuse.de- Commit a42b3d3 by Danny Sauer dsauerAATTsuse.com Fix broken LDAP connector test case syntax Reduce duplication in expected json method More refactoring, extended testing
* Move all JSON assembly logic into `expected_dex_json` definition
* This simplfied the examples so they were identical, so wrap them in a loop
* Using the loop allowed testing all four combinations of anon/not and starttls/ssl
* Test names look like `with Dex LDAP connectors with starttls has dex LDAP connectors for anon_admin`
* Wed Oct 03 2018 containers-bugownerAATTsuse.de- Commit f05863b by Daniel Orf dorfAATTsuse.com [bsc#1109025] LDAP connector can\'t be saved if anonymous authentication is selected
* Thu Sep 27 2018 containers-bugownerAATTsuse.de- Commit c16eda7 by Maximilian Meister mmeisterAATTsuse.de dont allow updating admin again when cluster update failed bsc#1108740 Signed-off-by: Maximilian Meister
* Wed Sep 26 2018 containers-bugownerAATTsuse.de- Commit 4a672de by Daniel Orf dorfAATTsuse.com [bsc#1109320] Fix external auth group mapping for group attr name
* Tue Sep 25 2018 containers-bugownerAATTsuse.de- Commit 31687a3 by Vicente Zepeda Mas vzepedamasAATTsuse.com FIX bsc#1099985 removes unassigned nodes and add a reject node link
* Tue Sep 18 2018 containers-bugownerAATTsuse.de- Commit 0c6f37f by Maximilian Meister mmeisterAATTsuse.de rebase patch to alter default time column value: 0_set_default_salt_events_alter_time_column_value.rpm.patch Signed-off-by: Maximilian Meister
* Mon Sep 17 2018 containers-bugownerAATTsuse.de- Commit 23bb090 by Maximilian Meister mmeisterAATTsuse.de use online status during update for update status we can make use of the online status to have a more accurate status during the update if a node is offline during the update, it likely means that it\'s rebooting Signed-off-by: Maximilian Meister
* Fri Sep 14 2018 containers-bugownerAATTsuse.de- Commit ce14443 by James Mason jmasonAATTsuse.com [BSC#1107545] Migrate subnet pillar name v2.1 -> v3
* Tue Sep 11 2018 containers-bugownerAATTsuse.de- Commit 88c5ba3 by Vítor Avelino vavelinoAATTsuse.com ldap (specs): minor improvements using capybara convetions Signed-off-by: Vítor Avelino
* Mon Sep 10 2018 containers-bugownerAATTsuse.de- Commit 967df9b by David Helkowski dhelkowskiAATTsuse.com Prevent changed external ldap values from skipping checks Alter external ldap auth form to redisable save and rerequire \'check connection\' when values are changed. (fate#324601)
* Fri Sep 07 2018 containers-bugownerAATTsuse.de- Commit 98b8e10 by David Helkowski dhelkowskiAATTsuse.com Fix test connection button for ldap external auth Corrected fetching of password for ldap external auth Corrected grabbing of certificate content for validating ldap external auth
* Fri Sep 07 2018 containers-bugownerAATTsuse.de- Commit 3d00130 by Maximilian Meister mmeisterAATTsuse.de update README fixes #643 Signed-off-by: Maximilian Meister
* Thu Sep 06 2018 containers-bugownerAATTsuse.de- Commit 68f23f5 by Rafael Fernández López ereslibreAATTereslibre.es Allow to configure Kubernetes log level globally. Also, include PSP and DNS addons. While they won\'t probably be modified, they could be used for certain use cases. Fixes: bsc#1106193
* Tue Sep 04 2018 containers-bugownerAATTsuse.de- Commit 1c3873f by Vítor Avelino contactAATTvitoravelino.me specs: fixed TempFile usage for file fixtures In some randomized seeds a temp file was being removed and no content was being uploaded, since the file didn\'t exist, and the file content wasn\'t being displayed on the screen as expected. Since there\'s no way of controlling when GC is going to remove the temp files, the only way is trying to keep the reference on hold to be usable. The solution for this was to simply return the TempFile instance from `to_file_fixture` instead of the path or base name. Signed-off-by: Vítor Avelino
* Wed Aug 29 2018 containers-bugownerAATTsuse.de- Commit 34c7b11 by Maximilian Meister mmeisterAATTsuse.de when rebooting admin, set tx_update_reboot_needed to false otherwise we have a window after the restart, where the reboot admin button is still shown, as the event proccessor still needs time to update the grains bsc#1099015 Signed-off-by: Maximilian Meister
* Wed Aug 29 2018 containers-bugownerAATTsuse.de- Commit 43d8e3d by James Mason jmasonAATTsuse.com Update bootstrap-slider bootstrap-slider >=10.1.0 includes https://github.com/seiyria/bootstrap-slider/pull/836 which was included as a patch on release-2.1. fix bsc#1107495
* Wed Aug 29 2018 containers-bugownerAATTsuse.de- Commit 65fc2e0 by Rafael Fernández López ereslibreAATTereslibre.es Update the minion hostname in a regular basis, based on the grain information. Fixes: bsc#1103307
* Wed Aug 29 2018 containers-bugownerAATTsuse.de- Commit af6595b by Joachim Gleissner jgleissnerAATTsuse.com Re-record cassettes required for cloud_cluster_spec Required due to change in salt API calls in app/models/cloud_cluster.rb. Commit fa09369 by Joachim Gleissner jgleissnerAATTsuse.com Add explanatory comment for sleep(1) work-around Add comment in app/models/cloud_cluster.rb that explains why there is currently a work-around in the form of sleep(1) required. Commit d2bb545 by Joachim Gleissner jgleissnerAATTsuse.com Sleep after calling saltutil.refresh_pillar Added sleep(1) after calling saltutil.refresh_pillar on the admin node when creating a cluster in the public cloud. Otherwise, the salt-cloud call that follows directly afterwards may still use old pillar data. Commit ae28c04 by Joachim Gleissner jgleissnerAATTsuse.com Call saltutil.sync_pillar runner before bootstrapping cloud cluster Added call_runner function to salt interface and added sync_pillar call to cloud_cluster.rb. This is required to activate the external pillar, which is normally done at orchestration time, but for creating cluster nodes in the cloud, the pillar is needed before that. See also https://github.com/kubic-project/velum/issues/399 for detailts. fixes bsc#1107082
* Wed Aug 29 2018 containers-bugownerAATTsuse.de- Commit b445cac by Marek Counts marek.countsAATToutlook.com Updated tool tips and removed default data - Added a database migration that removed default data from the database - Added smart documentation on each LDAP field in place of the default data. updates to the rspec tests too account for missing feilds due to the lack of defaults removed not required element ids removed not required element ids and corrected accidental removal of AATTcertificate_holder.port Removal of not needed fill_ins Under a previous assumption of the workings of the save button I had filled in stuff that is not needed for the test. Updated slim html to use Certificate Also updated unit tests to pass. Removal of extra line Corrected spelling and fixed spacing Corrected spelling and updated spacing to be consistent in the examples. update to tests to fix a bug There are some issues with these test causing random failures, Currently the best fix to this is locking down the run order. fixed rubocop complaints updated to fix rubocop complaints.
* Tue Aug 28 2018 containers-bugownerAATTsuse.de- Commit 0d201e7 by Marek Counts mcountsAATTsuse.com Updates to prevent auto complete from filling out LDAP fields A extra text field was added to break up the form flow to keep firefox from auto filling in the user name and password. this extra field was then hidden by css. if hidden in the HTML firefox would know and still auto fill.
* Mon Aug 27 2018 containers-bugownerAATTsuse.de- Commit 68e42c9 by Maximilian Meister mmeisterAATTsuse.de require uglifier only during packaging the constant `Uglifier` is not accessible when opening a rails console or running a rake command fix#uglifier Signed-off-by: Maximilian Meister
* Thu Aug 23 2018 containers-bugownerAATTsuse.de- Commit c4a814d by Maximilian Meister mmeisterAATTsuse.de prevent accessing discovery after bootstrap PR #627 introduced a weird behaviour of skip_before_action when \'only\' and \'if\' is used at the same time see: https://github.com/rails/rails/issues/9703 we can just omit the \'if setup_done?\' check because that\'s already implicitly happening through the application_controller Signed-off-by: Maximilian Meister
* Mon Aug 20 2018 containers-bugownerAATTsuse.de- Commit 377e4fe by James Mason jmasonAATTsuse.com PublicCloud#remove: Not yet implemented Removing nodes is not implemented in public clouds (Azure, EC2, GCE) at this time, so the interfaces for doing so should be disabled until the feature is implemented. fixes bsc#1107081
* Mon Aug 20 2018 containers-bugownerAATTsuse.de- Commit e3fa613 by Maximilian Meister mmeisterAATTsuse.de fix uglifier instantiation Signed-off-by: Maximilian Meister
* Tue Aug 14 2018 containers-bugownerAATTsuse.de- Commit d4648f9 by James Mason jmasonAATTsuse.com bsc#1098085 Add nodes in public clouds Requires #595 Requires #610 Adapt public cloud bootstrap views & controller actions to enable adding nodes after boostrap.
* Fri Aug 10 2018 containers-bugownerAATTsuse.de- Commit 3f08d4e by Daniel Orf dorfAATTsuse.com Added LDAP Connector validation button and logic Added new controller \"ldap_test_controller\". Updated _form.html.slim to add test button and message. Added ldap-conn-test.js to javascripts/settings. ldap-conn-test.js adds jquery logic to test button in place and modify message. Updated routes.rb for new controller. Added RSpec tests. Fixed Code Climate warnings. Added logic in ldap_test_controller_spec.rb to turn on/off based on LDAP server presence. Addressed comments from David Helkowski. Addressed comments from Vítor Avelino (x2). Updated \"Save\" button ID in dex_connector_ldap_feature_spec.rb. Added ldap_mock.rb, ldap_mock.yml, updated settings.scss (fate#324601)
* Fri Aug 10 2018 containers-bugownerAATTsuse.de- Commit 0b93b79 by Flavio Castelli fcastelliAATTsuse.com Fix smally typo This fixes issue #612 github-issue#612 Signed-off-by: Flavio Castelli
* Thu Aug 09 2018 containers-bugownerAATTsuse.de- Commit b38fb1f by Florian Bergmann fbergmannAATTsuse.de Make dex connector for LDAP use the file-based certificate upload. (fate#324601) Commit 0847881 by Florian Bergmann fbergmannAATTsuse.de Ensure that certificates are always associated with the new holder. Otherwise a reused certificate will not be associated with a new service. Commit 0ab1f88 by Florian Bergmann fbergmannAATTsuse.de Fix bsc1097753: Allow the user to upload a certificate via file. Commit 09c5822 by Florian Bergmann fbergmannAATTsuse.de Setup RSpec fixture path.
* Wed Aug 08 2018 containers-bugownerAATTsuse.de- Commit 830e027 by Vítor Avelino vavelinoAATTsuse.com ui: added openstack fields info and required validation The OpenStack API fields didn\'t have any additional information about them and no indicator of which fields were required or not for a valid configuration. Fixes bsc#1097817 Signed-off-by: Vítor Avelino
* Tue Aug 07 2018 containers-bugownerAATTsuse.de- Commit cd5cc07 by Maximilian Meister mmeisterAATTsuse.de keep information about adding nodes to the cluster after the bootstrap all the useful hints regarding autoyast were not accessible anymore bsc#1098085 Signed-off-by: Maximilian Meister
* Tue Aug 07 2018 containers-bugownerAATTsuse.de- Commit 388521f by Michal Jura mjuraAATTsuse.com [CPI] Split logic for Project and Domain ID, bsc#1103873
* Mon Aug 06 2018 containers-bugownerAATTsuse.de- Commit 4c32ca3 by Michal Jura mjuraAATTsuse.com Change CPI OpenStack Cinder text field to boolean button
* Mon Aug 06 2018 containers-bugownerAATTsuse.de- Commit e87f9ba by Jordi Massaguer Pla jmassaguerplaAATTsuse.de fix tests that were broken because PR#595 wasn\'t rebased before merging it. Signed-off-by: Jordi Massaguer Pla
* Mon Jul 30 2018 containers-bugownerAATTsuse.de- Commit f8d65aa by James Mason jmasonAATTsuse.com Add support for bootstrapping in Google Compute fixes bsc#1107080
* Tue Jul 24 2018 containers-bugownerAATTsuse.de- Commit 971c792 by Ty Daines tdainesAATTsuse.com Add ability to add, modify, delete and list LDAP connections in the ui (fate#324601)
* Thu Jul 19 2018 containers-bugownerAATTsuse.de- Commit d88991f by Michal Jura mjuraAATTsuse.com [CPI] Add option to ignore OpenStack Cinder availability zone, bsc#1095572 Ignore OpenStack Cinder avability zone when attaching volumes. When Nova and Cinder have different availability zones, this should be set to true. Default is false.
* Fri Jul 06 2018 containers-bugownerAATTsuse.de- Commit 52e869d by Maximilian Meister mmeisterAATTsuse.de rename migration file it needs to match the schema revision feature#external-ldap (fate#324601) Signed-off-by: Maximilian Meister Commit 9a9393c by David Helkowski dhelkowskiAATTsuse.com Add dex connector ldap table, models, and pillar output (fate#324601) Added a new table \"dex_connectors_ldap\" and migration file to the db schema and migrate files. Altered pillar controller to output this new data. Added a basic/empty app model for the new table so that the data can be fetched through it. Updated pillar rspec to test newly output connectors as well as to allow the empty connectors in the other pillar tests.
* Thu Jul 05 2018 containers-bugownerAATTsuse.de- Commit cd431fe by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1097754: Verify the a certificate is a valid X509 certificate. Commit 0ef31b8 by Florian Bergmann fbergmannAATTsuse.de Backport Rails 5 file_fixture method to access a fixture in a test.
* Thu Jul 05 2018 containers-bugownerAATTsuse.de- Commit b18c46e by Vítor Avelino vavelinoAATTsuse.com ui: renamed new nodes -> unassigned nodes On the cluster status summary we decided to rename \"new\" by \"unassigned\" to avoid mixing terms that may confuse the user. bsc#1100113 Signed-off-by: Vítor Avelino
* Mon Jul 02 2018 containers-bugownerAATTsuse.de- Commit f95f065 by Maximilian Meister mmeisterAATTsuse.de migrate once more after db creation this should enable CI to create newly added db tables/changes during a devenv deployment bsc#1098436 Signed-off-by: Maximilian Meister Commit b3434c3 by Maximilian Meister mmeisterAATTsuse.de save online status through reconciler (bsc#1098436) when a minion is down, it should be shown as down in the UI separate columns in the UI for applied and online status Signed-off-by: Maximilian Meister
* Thu Jun 28 2018 containers-bugownerAATTsuse.de- Commit 1e98beb by Vítor Avelino vavelinoAATTsuse.com ui: removed fixed width for main header logo Logo can be both CaaSP or Kubic now and the fixed width was causing a small misalignment with Kubic logo. Turns out that this fixed width is not necessary at all. fix#kubic-logo Signed-off-by: Vítor Avelino
* Thu Jun 28 2018 containers-bugownerAATTsuse.de- Commit 713539d by Maximilian Meister mmeisterAATTsuse.de add branding images before precompilation a simple mount through the manifests is not enough images need to be precompiled velum#branding Signed-off-by: Maximilian Meister
* Fri Jun 22 2018 containers-bugownerAATTsuse.de- Commit f94cdfd by Jordi Massaguer Pla jmassaguerplaAATTsuse.de update sprockets to fix cve-2018-3760 cve#2018-3760 Signed-off-by: Jordi Massaguer Pla
* Thu Jun 21 2018 containers-bugownerAATTsuse.de- Commit 4303f26 by Florian Bergmann fbergmannAATTsuse.de Load variables used to render autoyast in separate class. Commit 875ac30 by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1072242: Use keyboard layout from admin installation. The keyboard layout the user was using during the installation will be used in the autoyast generated to bootstrap additional nodes. Signed-off-by: Florian Bergmann
* Wed Jun 20 2018 containers-bugownerAATTsuse.de- Commit e943e07 by Vítor Avelino vavelinoAATTsuse.com ui: don\'t show failed alert for repeated orchestration A failed alert is shown whenever one of the nodes fails to bootstrap. This is an optimization to only show this alert once for a specific orchestration run. If the user closes the alert, it won\'t come back unless a new orchestration happens. bsc#1097752 Signed-off-by: Vítor Avelino
* Thu Jun 14 2018 containers-buildsAATTsuse.de- Commit 62237be by Flavio Castelli fcastelliAATTsuse.com Expose pillars managing kubernetes extra args The salt backend already has pillars that can be used to provide extra arguments to the different kubernetes components (api-server, controller-manager, scheduler, kubelet and proxy). This commit allows to provide values for these settings. This is something we should
*
*not
*
* expose into the UI. It can be handy with certain customers who are looking for a way to pass not-yet-handled flags to kubernetes. Signed-off-by: Flavio Castelli feature#hidden-kubernetes-pillars
* Thu Jun 14 2018 containers-buildsAATTsuse.de- Commit 0690ea7 by James Mason jmasonAATTsuse.com Add a feature spec for building cluster in Azure Most of the content is covered in the cloud_cluster model, but there are a few differentiators that should be checked.
* Wed Jun 13 2018 containers-buildsAATTsuse.de- Commit 554d1f2 by Maximilian Meister mmeisterAATTsuse.de make 503 page more generic bsc#1080636 Signed-off-by: Maximilian Meister
* Tue Jun 12 2018 containers-buildsAATTsuse.de- Commit 0f081da by Jordi Massaguer Pla jmassaguerplaAATTsuse.de fix favicon path Commit 4f4d1e9 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de fix message in 503 Commit f267cec by Jordi Massaguer Pla jmassaguerplaAATTsuse.de move branding images to branding dir Signed-off-by: Jordi Massaguer Pla Commit f290a30 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de move branding to public/branding Commit 1706def by Jordi Massaguer Pla jmassaguerplaAATTsuse.de Make 503 error message \"branding agnostic\" This way this error message can be used both in kubic and in caasp This is needed because we need to revert a change in the velum-branding package and the caasp-container-manifests that was mounting the public directory from the host. Mounting the public directory from the host, was \"hidding\" the assets files present in the public directory and so our images were broken. This was only happening in the iso and not in our dev environments because in production we have the assets precompiled and we don\'t compile them on the fly, since we don\'t have a js engine installed. Signed-off-by: Jordi Massaguer Pla
* Mon Jun 11 2018 containers-buildsAATTsuse.de- Commit 4d19ba2 by Rafael Fernández López ereslibreAATTereslibre.es Serve the flannel backend on the Pillar API If we don\'t serve the flannel backend on the Pillar API, the upgrade will use the default in the `sls` files, that is: `vxlan`. The intent was always to keep `udp` for upgraded systems until we migrate them to the `vxlan` backend, so make the Pillar controller serve the current `flannel:backend` pillar value that is present on the database (it will be `vxlan` for new installations, and `udp` for systems coming from the 2.0 release). Fixes: bsc#1096710
* Tue Jun 05 2018 containers-buildsAATTsuse.de- Commit 1116260 by James Mason jmasonAATTsuse.com Change cloud cluster view based on cluster size
* If existing nodes are in progress/built; say so.
* If we\'ve already met the 3-node-minimum, no need for the tip Commit 0f48c32 by James Mason jmasonAATTsuse.com Implement a decorator for cluster size slider Slider range is dynamically adjusted by decorated class Commit 18f4184 by James Mason jmasonAATTsuse.com Make CloudCluster model aware of cluster size limits Commit 8c1eeed by James Mason jmasonAATTsuse.com Update assets for bootstrap-slider Including https://github.com/seiyria/bootstrap-slider/pull/836 fixes: bsc#1093565
* Thu May 31 2018 containers-bugownerAATTsuse.de- Commit f50fb8a by Vítor Avelino vavelinoAATTsuse.com ui: fix forced node removal modal paragraphs bsc#1095252 Signed-off-by: Vítor Avelino
* Thu May 31 2018 containers-bugownerAATTsuse.de- Commit 7e70efd by Vítor Avelino vavelinoAATTsuse.com ui: show proper update failed message and reboot button Changed update failed error message stating a generic state instead of saying that the update failed itself. Update admin node link only showing if `tx_update_reboot_needed` flag is truthy. Otherwise keeps it hidden. bsc#1082786 Signed-off-by: Vítor Avelino
* Wed May 30 2018 containers-bugownerAATTsuse.de- Commit 8fcb730 by Miquel Sabaté Solà msabateAATTsuse.com reservations: don\'t save if a value is not valid Before this commit, the transaction on the `create` method from the KubeletComputeResourcesReservationsController was useless because it updated the attributes before the transaction itself. This commit corrects that and simplifies it, so an error on any of the parameters will issue a rollback (hence, the DB will be left untouched). Signed-off-by: Miquel Sabaté Solà Commit be1c78a by Miquel Sabaté Solà msabateAATTsuse.com pillar: added a validator for hard eviction The kubelet component does not accept any kind of value: it expects a specific syntax. This commit adds a validator for the pillar that applies only if its `pillar` value is set to `kubelet:eviction-hard`. bsc#1093894 Signed-off-by: Miquel Sabaté Solà
* Wed May 30 2018 containers-bugownerAATTsuse.de- Commit f3a38e6 by Maximilian Meister mmeisterAATTsuse.de read distribution from a file this enables branding for kubic and caasp feature#branding Signed-off-by: Maximilian Meister
* Tue May 29 2018 containers-bugownerAATTsuse.de- Commit 96437a9 by Flavio Castelli fcastelliAATTsuse.com Revert \"Allow users to choose flannel\'s backend\" This reverts commit 99550aacb6b7c5bfa48545005039c743aa2ee853. It turns out it\'s possible to use the VXLAN backend to create clusters over segmented networks. feature#revert-UDP-selection-UI
* Tue May 29 2018 containers-bugownerAATTsuse.de- Commit 5334257 by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1062186: update tooltip to explain possible dashboard values.
* Mon May 28 2018 containers-bugownerAATTsuse.de- Commit 5463971 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de update velum requirements on mariadb for sle15 Signed-off-by: Jordi Massaguer Pla
* Fri May 25 2018 containers-bugownerAATTsuse.de- Commit a93be89 by David Cassany dcassanyAATTsuse.com Make use of %license macro
* Fri May 25 2018 containers-bugownerAATTsuse.de- Commit 7e6fdf1 by Vítor Avelino vavelinoAATTsuse.com ux: toggle domain/project id/name inputs Whenever the user wants to configure openstack domain/project id/name, they cannot fill both cases. To prevent that we are disabling the respective opposite fields when one of the fields is filled. bsc#1091809 ui improvement Signed-off-by: Vítor Avelino
* Thu May 24 2018 containers-bugownerAATTsuse.de- Commit b04fc47 by David Cassany dcassanyAATTsuse.com Spec cleaning - Removing systemd dependency - Removing duplicated Provides
* Thu May 24 2018 containers-bugownerAATTsuse.de- Commit f7d227d by Maximilian Meister mmeisterAATTsuse.de show apply button after each setting creation it makes sense to offer the apply button directly after a user has created a new setting otherwise you have to navigate back to the setting index which is not a good UX because it\'s hidden in many cases a user just wants to create one setting and then immediately apply settings#apply-on-create Signed-off-by: Maximilian Meister
* Wed May 23 2018 containers-bugownerAATTsuse.de- Commit 456c04d by tdaines42 tdainesAATTsuse.com made changes so project id and domain id can be used for openstack
* Wed May 23 2018 containers-bugownerAATTsuse.de- Commit 7ef3502 by Vítor Avelino vavelinoAATTsuse.com ui: fixed unwanted acceptance failure alert Also replaced deprecated ajax callback method. bsc#1093869 follow-up Signed-off-by: Vítor Avelino
* Tue May 22 2018 containers-bugownerAATTsuse.de- Commit 59f3741 by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1090067 (5/5): Extract more logic into the base controller. The base controller for certificates now also handles the `new`, `edit`, `update` and `destroy` actions. Only the registry_mirror keeps handling its own `create` as it accesses the `registry` object as well. (cherry picked from commit 16ee911dfa89e84e5366886ebc00a8afffb82c86) Commit 01ed07e by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1090067 (4/5): Move the migration to the end of list. The tables for system_certificates were not part of the /var/lib/velum/schema.rb file and were not migrated. (cherry picked from commit 471f9eda08ad99f56d0d03cf002a308674fb27fa) Commit 25dc6c9 by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1090067 (3/5): UI for self-signed certificates on welcome page. Rendering of the system form in the welcome page uses a partial. (cherry picked from commit 61b72b5f4b4cebd62e9f31e299d07e068f7c0270) Commit 22af67c by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1090067 (2/5): Extract certificate handling logic. (cherry picked from commit 5e85427dfaa3db1795d1cca92440d334e6dcd182) Commit ad18b17 by Florian Bergmann fbergmannAATTsuse.de Fix bsc#1090067 (1/5): implementation of a system wide certificates. The implementation follows the registry & registry mirror style and adds a simple (name, certificate) model.
* Tue May 22 2018 containers-bugownerAATTsuse.de- Commit a660c5b by Vítor Avelino vavelinoAATTsuse.com js: removal of pending acceptance state We are storing the minion id in session storage while it goes under pending acceptance state. However we were not handling its removal if the acceptance succeeded or if an error happened after the request. This patch fixes those scenarios avoiding unwanted user experiences. bsc#1093869 Signed-off-by: Vítor Avelino
* Mon May 21 2018 containers-bugownerAATTsuse.de- Commit 5312422 by Vítor Avelino vavelinoAATTsuse.com ui: show kubectl config download link on auth page Some people that are not the admin user need the kubectl config file. This link is only accessible in the dashboard page. This patch shows s a link on for anonymous users on the authentication page. bsc#1093528 Signed-off-by: Vítor Avelino
* Mon May 21 2018 containers-bugownerAATTsuse.de- Commit dc2884b by Flavio Castelli fcastelliAATTsuse.com Allow override of CNI related pillars Make Velum aware of some pillars related with CNI/cilium. That makes possible to switch to cilium. No UI is needed for that right now. feature#cilium-selected-preview-on-v3
* Sat May 19 2018 containers-bugownerAATTsuse.de- Commit 7af7b29 by Flavio Castelli fcastelliAATTsuse.com Provide UI to manage kubernetes auditing feature#auditing fate#325337
* Fri May 18 2018 containers-bugownerAATTsuse.de- Commit 9c31386 by Florian Bergmann fbergmannAATTsuse.de Redirect to registry page instead of mirror when editing a registry.
* Thu May 17 2018 containers-bugownerAATTsuse.de- Commit 4e4eaff by Maximilian Meister mmeisterAATTsuse.de drop some css that overwrites bootstrap button defaults ui#css-enhancements Signed-off-by: Maximilian Meister Commit b3251e6 by Maximilian Meister mmeisterAATTsuse.de emphasize settings apply workflow the warning that any changes are not immediately applied should be more visible a blue flash message background with a blue button is not visible enough settings-ui#apply Signed-off-by: Maximilian Meister
* Thu May 17 2018 containers-bugownerAATTsuse.de- Commit 99550aa by Flavio Castelli fcastelliAATTsuse.com Allow users to choose flannel\'s backend Starting with v3 we changed the default flannel backend from UDP to VXLAN. The VXLAN backend does not work when users have segmented networks, while the UDP one does. The salt states are already capable of switching between the two backends, we just needed to expose the right pillars from Velum. This commit changes the setup UI to allow users to choose the backend of and the port used by flannel. feature#flannel-backend Signed-off-by: Flavio Castelli
* Wed May 16 2018 containers-bugownerAATTsuse.de- Commit 95ecd9e by Rafael Fernández López ereslibreAATTereslibre.es Mark admin as updated when we are going to reboot it This ensures that when the admin is back, the status of the minion on the database will match the grains from the very beginning, without having to rely on the background task updating all minions (including the admin) update status. Fixes: bsc#1092910
* Tue May 15 2018 containers-bugownerAATTsuse.de- Commit 546b96a by Maximilian Meister mmeisterAATTsuse.de validate pillars dont allow whitespaces at the beginning and at the end bsc#1087447 Signed-off-by: Maximilian Meister Commit 878dde3 by Maximilian Meister mmeisterAATTsuse.de dont thrown away errors when applying pillars bsc#1087447 Signed-off-by: Maximilian Meister
* Tue May 15 2018 containers-bugownerAATTsuse.de- Commit 3678b52 by Flavio Castelli fcastelliAATTsuse.com Handle kubelet resource reservations Allow users to specify the compute resource reservations for the `kube` and the `system` components. This is going to be done by using a new page available under the `settings` section. This is required to address bsc#1086185
* Tue May 15 2018 containers-bugownerAATTsuse.de- Commit 4ec7dba by Maximilian Meister mmeisterAATTsuse.de rebase patch it got wrongly rebased in #502 fix#patch Signed-off-by: Maximilian Meister
* Tue May 15 2018 containers-bugownerAATTsuse.de- Commit 8b75d9f by Kiall Mac Innes kiallAATTmacinnes.ie Add Housekeeping Job
* Fri May 11 2018 containers-bugownerAATTsuse.de- Commit 9d97b1d by James Mason jmasonAATTsuse.com Prevent double-clicking setup#bootstrap submit in public cloud Since that actually triggered construction of nodes, we need just a little something to keep the impatient from clicking over & over.
* Wed May 09 2018 containers-bugownerAATTsuse.de- Commit 9d97b1d by James Mason jmasonAATTsuse.com Prevent double-clicking setup#bootstrap submit in public cloud Since that actually triggered construction of nodes, we need just a little something to keep the impatient from clicking over & over.
* Wed May 09 2018 containers-bugownerAATTsuse.de- Commit f24a3f1 by James Mason jmasonAATTsuse.com Include counts on cloud bootstrap jobs in discovery UI Commit f50a986 by James Mason jmasonAATTsuse.com Return cloud bootstrapping job progress in discovery BONUS: render discovery.html faster by not evaluating unused hash conditions. Commit 1c00b7b by James Mason jmasonAATTsuse.com Move cloud cluster building into model; capture job ids... ... clear failures on (re)start Commit 8fcedc8 by James Mason jmasonAATTsuse.com Add an event handler for completing cloud bootstrap Commit d50aad9 by James Mason jmasonAATTsuse.com Define a model for persisting salt job progress.
* Wed May 09 2018 containers-bugownerAATTsuse.de- Commit 1f3adaa by Rafael Fernández López ereslibreAATTereslibre.es Make the `bin/init` process only do initialization operations Do not make this script run the puma server. This way, this command can be run as an init container, leaving the server execution to the pod definition. Fixes: bsc#1091843
* Wed May 09 2018 containers-bugownerAATTsuse.de- Commit 45bee11 by Rafael Fernández López ereslibreAATTereslibre.es Use `update_columns` to update minion attributes with the minion reconciler This will avoid strange issues when Rails caches the attributes of the `minions` table (we also avoid the need to `reload` explicitly for this corner cases). Fixes: bsc#1091843
* Tue May 08 2018 containers-bugownerAATTsuse.de- Commit add62fa by Vítor Avelino vavelinoAATTsuse.com Fixed misleading text about nodes selection Small fix of a misleading text that referred to an old behavior of the nodes discovery page. bsc#1059787 Signed-off-by: Vítor Avelino
* Tue May 08 2018 containers-bugownerAATTsuse.de- Commit 8ef0c57 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de update version to 4.0.0+dev
* Fri Apr 27 2018 containers-bugownerAATTsuse.de- Commit 162c0d0 by Liam Proven lprovenAATTsuse.com Reword force-removal warning as per IRC #docker discussion
* Fri Apr 27 2018 containers-bugownerAATTsuse.de- Commit 01fe462 by MalloZup dmaiocchiAATTsuse.com change link to actual one
* Thu Apr 26 2018 containers-bugownerAATTsuse.de- Commit 310d614 by Vítor Avelino vavelinoAATTsuse.com node removal: improved way of showing removal links \'Remove\' and \'Force remove\' links were being shown at the same time so far. But \'Force remove\' only makes sense to be shown when a previous removal has failed. With this we are showing only one link at a time. When the highstate of a minion is fine, we show \'Remove\'. Whenever it comes \'removal_failed\', we show \'Force remove\'. enhancemente#node-removal Signed-off-by: Vítor Avelino
* Fri Apr 20 2018 containers-bugownerAATTsuse.de- Commit 310d614 by Vítor Avelino vavelinoAATTsuse.com node removal: improved way of showing removal links \'Remove\' and \'Force remove\' links were being shown at the same time so far. But \'Force remove\' only makes sense to be shown when a previous removal has failed. With this we are showing only one link at a time. When the highstate of a minion is fine, we show \'Remove\'. Whenever it comes \'removal_failed\', we show \'Force remove\'. enhancemente#node-removal Signed-off-by: Vítor Avelino
* Fri Apr 20 2018 containers-bugownerAATTsuse.de- Commit 514916c by Flavio Castelli fcastelliAATTsuse.com Rename pillar used to choose the CRI The pillar name has been changed by latest commits to the salt repository. feature#crio Signed-off-by: Flavio Castelli Commit 3864cce by Flavio Castelli fcastelliAATTsuse.com Bootstrap settings: cri-o support Allow the user to choose between different container runtimes. This enables users to create clusters backed by cri-o instead of docker. feature#crio Signed-off-by: Flavio Castelli Co-authored-by: Vítor Avelino
* Thu Apr 19 2018 containers-bugownerAATTsuse.de- Commit 3ce16bd by Rafael Fernández López ereslibreAATTereslibre.es Fix rubocop offenses Feature#clean-code
* Thu Apr 19 2018 containers-bugownerAATTsuse.de- Commit 569db72 by Flavio Castelli fcastelliAATTsuse.com Support the kubernetes feature gates pillar Allow users to override the kubernetes:feature_gates pillar via our external pillar source. feature#feature-gates Signed-off-by: Flavio Castelli
* Thu Apr 19 2018 containers-bugownerAATTsuse.de- Commit 7325d5f by Vítor Avelino vavelinoAATTsuse.com Introduce force removal feature
* Do not poll for grains from the UI Remove all polling to the salt-api triggered from the UI, this will make the UI responsive when some minion is down. We\'ll update the relevant grains for us in the background.
* Add the admin node as part of the list of minions in the database It allows us to keep track of the update flags of the admin node
* Include a new scope `cluster_role` for minions that will return the minions that are part of the Kubernetes cluster (not admin nodes) Fixes: bsc#1085291 feature#force-node-removal
* Mon Apr 16 2018 containers-bugownerAATTsuse.de- Commit e46dca7 by Michal Jura mjuraAATTsuse.com Skip worker_bootstrap when cloud_framework is openstack, bsc#1089652 We don\'t provide openstack-types.json file for OpenStack cloud framework. We have to skip worker_bootstrap function from SetupController class.
* Thu Apr 12 2018 containers-bugownerAATTsuse.de- Commit 3364855 by Kiall Mac Innes kiallAATTmacinnes.ie Add and run a import_pillar_seeds rake task This is a generic rake task that will import seed pillar values from all files in /etc/caasp/pillar-seeds/. The format of files in this folder should match: [#] cat /etc/caasp/pillar-seeds/test.yaml - pillar: cloud:framework value: openstack feature#hide-openstack-cpi
* Thu Apr 12 2018 containers-bugownerAATTsuse.de- Commit bd18f91 by Rafael Fernández López ereslibreAATTereslibre.es Avoid the event processor from crashing if it cannot interpret the event arguments If we executed: `salt-run state.orchestrate 2` we get an error that crashes the event processor: `TypeError: no implicit conversion of String into Integer` This happens because of `fun_args.first[\"mods\"]` being `2[\"mods\"]`, what makes it crash with the previous exception. `2.respond_to?(:[])` is `true`, so I think the best thing we can do here is to swallow the exception, and ignore the event if we cannot even process its arguments. Fixes: bsc#1088597
* Mon Apr 09 2018 containers-bugownerAATTsuse.de- Commit 349e38b by James Mason jmasonAATTsuse.com Add cursor: pointer to .instance-type-box re: https://github.com/kubic-project/velum/pull/488#issuecomment-378876002 re: https://github.com/kubic-project/velum/pull/449#pullrequestreview-101140932 Commit 47044f0 by James Mason jmasonAATTsuse.com PubCloud#Azure Add a help link for authentication inputs Commit 5b175f3 by James Mason jmasonAATTsuse.com PubCloud#UI Resize input fields on public cloud bootstrap ... based on actual browser testing Commit c1e842c by James Mason jmasonAATTsuse.com PubCloud#Azure Add an input when bootrapping on Azure The storage_account attribute was discovered to be required, not optional.
* Mon Apr 09 2018 containers-bugownerAATTsuse.de- Commit ee7e6f1 by Vítor Avelino contactAATTvitoravelino.me Fixed # of masters warning when adding a new master This patch fixes the warning that was being shown when the user selected a new master node to be added and it wasn\'t breaking the odd number constraint. Before this, the current assigned masters were not being considered to show/not show the warning. Now the warning will only be shown if the number of masters to be added breaks the constraint as it\'s expected to be. bsc#1087267 Signed-off-by: Vítor Avelino
* Fri Apr 06 2018 containers-bugownerAATTsuse.de- Commit 6171165 by Vítor Avelino contactAATTvitoravelino.me Fixed # of masters warning when removing worker This patch fixes the warning modal that was being shown when the cluster was in an unsupported configuration (even # masters) and user was trying to remove a normal worker node. This scenario should trigger that warning modal. bsc#1087123 Signed-off-by: Vítor Avelino
* Fri Apr 06 2018 containers-bugownerAATTsuse.de- Commit 03c5c49 by Vítor Avelino contactAATTvitoravelino.me Hide CPI panel when no cloud framework When no cloud:framework pillar is set, it means that CaaSP is not running in a cloud environment. So it doesn\'t make sense to show that option in that scenario. feature#cpi Signed-off-by: Vítor Avelino
* Thu Apr 05 2018 containers-bugownerAATTsuse.de- Commit ff412c6 by James Mason jmasonAATTsuse.com Correct rubocop target ruby version Rubcop only considers major.minor: https://github.com/bbatsov/rubocop/issues/2663 Commit 81f7409 by James Mason jmasonAATTsuse.com Update preferred ruby dev version to match SLE12 release 2.1.2 -> 2.1.9
* Wed Apr 04 2018 containers-bugownerAATTsuse.de- Commit 3320a77 by Rafael Fernández López ereslibreAATTereslibre.es Properly create and update orchestrations that have been executed out of Velum It\'s more correct for the removal orchestration to read the parameters that were passed to the orchestration rather than to rely on the current status of the minions on the database.
* Tue Apr 03 2018 containers-bugownerAATTsuse.de- Commit 2fd6eee by Maximilian Meister mmeisterAATTsuse.de update rails-html-sanitizer (bsc#1086598) CVE-2018-3741 Signed-off-by: Maximilian Meister
* Mon Apr 02 2018 containers-bugownerAATTsuse.de- Commit 27645e8 by Vítor Avelino contactAATTvitoravelino.me Added alert in overview page when unsupported config Whenever the cluster goes to an unsupported configuration state, an alert is shown with the reason to keep the user aware of that. enhancement#unsupported_config_msg Signed-off-by: Vítor Avelino
* Mon Apr 02 2018 containers-bugownerAATTsuse.de- Commit f1277f2 by Vítor Avelino contactAATTvitoravelino.me Added minimum of three nodes to warning This is a follow up from the previous node removal feature patch that shows the requirement of minimum of three nodes when user removes a node. enhancement#follow_up_node_removal Signed-off-by: Vítor Avelino
* Mon Mar 26 2018 containers-bugownerAATTsuse.de- Commit 8a35b90 by Rafael Fernández López ereslibreAATTereslibre.es Migrate LDAP passwords Fixes: bsc#1071023
* Mon Mar 26 2018 containers-bugownerAATTsuse.de- Commit ec69c13 by James Mason jmasonAATTsuse.com Extend salt-api timeout as long as possible. Allow enough time for the salt timeout, and a minion timeout before cutting off the API call. Should resolve https://github.com/kubic-project/velum/issues/456 introduced in 6189bcad .
* Fri Mar 23 2018 containers-bugownerAATTsuse.de- Commit ccfc31c by Vítor Avelino vavelinoAATTsuse.com Add ability to remove nodes. This PR adds a `Remove` link to the interface, which will execute the `orch.removal` orchestration in salt with the `minion_id` of the selected node to be removed. Master nodes are not removable yet. feature#node_removal Signed-off-by: Vítor Avelino Co-authored-by: Rafael Fernández López
* Fri Mar 23 2018 containers-bugownerAATTsuse.de- Commit 184e46f by Michal Jura mjuraAATTsuse.com Run rake task and import Cloud Provider config file Commit 86ae24f by Michal Jura mjuraAATTsuse.com Add rake task to import CPI config
* Thu Mar 22 2018 containers-bugownerAATTsuse.de- Commit 1d2b129 by Vítor Avelino vavelinoAATTsuse.com Improved CPI behavior with cloud frameworks If a cloud framework is previously set, CPI value, if enabled, would be based on what was set. For example, if cloud:framework pillar value is \"ec2\", we would set \"ec2\" as the CPI value if enabled. The same be applied for GCE and Azure. Otherwise, \"openstack\" would be the default value for CPI if enabled. feature#cpi Signed-off-by: Vítor Avelino
* Thu Mar 22 2018 containers-bugownerAATTsuse.de- Commit 8f752e8 by Vítor Avelino contactAATTvitoravelino.me ui: allowed master node to be added to the cluster So far we only allowed workers to be added to the cluster through the unassigned nodes page. With this patch, we are now allowing the users to select the roles of the nodes they want to add to the cluster. Very similar to what is available from the discovery bootstrap page. This feature keep forcing the user to respect the contrainst of having an odd number of master nodes so we have a stable topology. feature#add_master_node Signed-off-by: Vítor Avelino
* Thu Mar 22 2018 containers-bugownerAATTsuse.de- Commit 7a13616 by Vítor Avelino contactAATTvitoravelino.me js: improved hostnames validation and workflow This is a follow-up of 8017753. After QA, this fixes an issue and improves the workflow of accepting pending nodes. Signed-off-by: Vítor Avelino
* Wed Mar 21 2018 containers-bugownerAATTsuse.de- Commit 5aa9cf5 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de update loofah to fix xss vulnerability cve#2018-8048 bsc#1085967 Signed-off-by: Jordi Massaguer Pla
* Wed Mar 21 2018 containers-bugownerAATTsuse.de- Commit e8d455d by Michal Jura mjuraAATTsuse.com Add cloud:provider to cloud_provider_contents for Salt API call
* Wed Mar 21 2018 containers-bugownerAATTsuse.de- Commit af4f939 by Maximilian Meister mmeisterAATTsuse.de typo fix#typo Signed-off-by: Maximilian Meister
* Thu Mar 15 2018 containers-bugownerAATTsuse.de- Commit 55ceb81 by Maximilian Meister mmeisterAATTsuse.de enable applying registry settings feature#registry Signed-off-by: Maximilian Meister Co-Authored-by: Vítor Avelino
* Wed Mar 14 2018 containers-bugownerAATTsuse.de- Commit 943655d by Vítor Avelino vavelinoAATTsuse.com Added CPI configuration During the boostrap configuration it\'s now possible to enable cloud provider integration. Co-authored-by: Maximilian Meister Signed-off-by: Vítor Avelino
* Mon Mar 12 2018 containers-bugownerAATTsuse.de- Commit 52980f3 by Kiall Mac Innes kiallAATTmacinnes.ie Fix db/seeds creation of SUSE registry URL record This currently fails, as the method does not exist: NoMethodError: undefined method `find_or_initialize\' for [#]
* Mon Mar 12 2018 containers-bugownerAATTsuse.de- Commit c6c9184 by Maximilian Meister mmeisterAATTsuse.de prevent multiple orchestrations at the same time Signed-off-by: Maximilian Meister
* Fri Mar 09 2018 containers-bugownerAATTsuse.de- Commit 9e1dbe6 by Vítor Avelino contactAATTvitoravelino.me Add settings page to configure registries feature#settings-page Signed-off-by: Vítor Avelino Co-Authored-By: Maximilian Meister
* Wed Mar 07 2018 containers-bugownerAATTsuse.de- Commit a9a139d by James Mason jmasonAATTsuse.com PubCloud#Azure Add support for Microsoft Azure framework Commit 3172f83 by James Mason jmasonAATTsuse.com Fix up a misnamed spec
* Tue Mar 06 2018 containers-bugownerAATTsuse.de- Commit 1003acb by Rafael Fernández López ereslibreAATTereslibre.es Rename \"Retry cluster upgrade\" to \"Retry cluster update\". I didn\'t refactor down the line because ultimately we cannot have an enum with a :update status (rails will complain, so at least the enum on the Orchestration model would need to be called :upgrade). For that reason, keep everything as it is (call internally `upgrade`, but just show \"Retry cluster update\" on the button label).
* Thu Mar 01 2018 containers-bugownerAATTsuse.de- Commit 1003acb by Rafael Fernández López ereslibreAATTereslibre.es Rename \"Retry cluster upgrade\" to \"Retry cluster update\". I didn\'t refactor down the line because ultimately we cannot have an enum with a :update status (rails will complain, so at least the enum on the Orchestration model would need to be called :upgrade). For that reason, keep everything as it is (call internally `upgrade`, but just show \"Retry cluster update\" on the button label).
* Tue Feb 27 2018 containers-bugownerAATTsuse.de- Commit b74718c by Maximilian Meister mmeisterAATTsuse.de disable codeclimate check that rubocop already does this is one example of overlapping checks that are performed by rubocop and therefore can be disabled, otherwise it shows up on the github status always Signed-off-by: Maximilian Meister
* Mon Feb 26 2018 containers-bugownerAATTsuse.de- Commit 2115e91 by Maximilian Meister mmeisterAATTsuse.de rebase patch Signed-off-by: Maximilian Meister
* Mon Feb 26 2018 containers-bugownerAATTsuse.de- Commit 50aa226 by Maximilian Meister mmeisterAATTsuse.de add mirrors model Signed-off-by: Maximilian Meister Commit d621057 by Maximilian Meister mmeisterAATTsuse.de drop docker prefix for registries Signed-off-by: Maximilian Meister
* Mon Feb 26 2018 containers-bugownerAATTsuse.de- Commit 6f880cf by Maximilian Meister mmeisterAATTsuse.de update a few gems we needed to have a new version of net-ldap because of CVE-2017-17718 this needed an update for devise_ldap_authenticatable Signed-off-by: Maximilian Meister
* Thu Feb 22 2018 containers-bugownerAATTsuse.de- Commit 4eb2d26 by James Mason jmasonAATTsuse.com Add user interface for public cloud bootstrapping new file: app/assets/stylesheets/pages/instance_type.scss
* Thu Feb 15 2018 containers-bugownerAATTsuse.de- Commit 6189bca by Kiall Mac Innes kiallAATTmacinnes.ie Reduce salt-api timeout from 60 sec, to 30 sec HAProxy has a 50 sec timeout, meaning HAProxy will kill the connection before Velum times out waiting for salt-api when there are issues.
* Thu Feb 08 2018 containers-bugownerAATTsuse.de- Commit b12fdf4 by Kiall Mac Innes kiallAATTmacinnes.ie Add manual kubeconfig setup instructions Add instructions for manually adding a cluster to a pre-existing kubeconfig file.
* Wed Feb 07 2018 containers-bugownerAATTsuse.de- Commit 2db24eb by Kiall Mac Innes kiallAATTmacinnes.ie Use separate Dex clients for each actual client Previously Velum, CaaSP CLI, and Kubernetes all shared a single Dex client. From a security perspective, this was far from ideal. Update Velum to:
* Generate a unique secret for the Velum and Kubernetes client during setup
* Add a migration to generate secrets during upgrade
* Use the Velum client to auth with Dex
* Request a token from Dex which is valid for the kubernetes client
* Tue Feb 06 2018 containers-bugownerAATTsuse.de- Commit 48188f1 by Chris Olstrom chrisAATTolstrom.com Add fallback link to fetch kubeconfig if redirect fails
* Tue Feb 06 2018 containers-bugownerAATTsuse.de- Commit 01453e3 by James Mason jmasonAATTsuse.com Update login feature spec I had issues with a failing test that used inconsitent access to the \'Log in\' button, so I\'ve updated it to match the rest of the spec. Additionally, the descriptions didn\'t make sense in documentation format so I reworded them.
* Fri Feb 02 2018 containers-bugownerAATTsuse.de- Commit b58ccc8 by Maximilian Meister mmeisterAATTsuse.de relax gems Signed-off-by: Maximilian Meister Commit a02e6b9 by Maximilian Meister mmeisterAATTsuse.de use bundler dependency script from darix also hardcode the native dependencies in the spec template as they likely won\'t change that often Signed-off-by: Maximilian Meister - patch modifications: + 1_nodev.gem.patch
* Thu Feb 01 2018 containers-bugownerAATTsuse.de- Commit 8017753 by Vítor Avelino contactAATTvitoravelino.me js: added unique hostnames validation Velum now validates if selected nodes to bootstrap and be added to the cluster have unique hostnames. The error message appears in an error alert.
* Thu Feb 01 2018 containers-bugownerAATTsuse.de- Commit 5a64a71 by Maximilian Meister mmeisterAATTsuse.de bump uglifier it got updated in factory Signed-off-by: Maximilian Meister
* Tue Jan 30 2018 containers-bugownerAATTsuse.de- Commit ab33601 by Rafael Fernández López ereslibreAATTereslibre.es Install `bundler` after `bundle clean`. `bundle clean` also removes stale executables under `bin`. It is fine that bundler thinks that `bundler` executable is stale, but it\'s not, we are just installing it as a separate gem, so do that at the end of the spec, after we have executed `bundle clean`.
* Mon Jan 29 2018 containers-bugownerAATTsuse.de- Commit 461cbf8 by Rafael Fernández López ereslibreAATTereslibre.es Add `assets` group that will only be loaded when we are compiling the assets
* Mon Jan 29 2018 containers-bugownerAATTsuse.de- Commit 3501c8b by Maximilian Meister mmeisterAATTsuse.de skip bundler during spec file creation a bundler BuildRequire is already hardcoded other than that it prevents a false strict dependency on bundler, e.g. if i run make_spec.sh on openSUSE 42.1 it will render bundler 1.10.6 whatever is installed on the system. this happened in our concourse build container that uses 42.1 as a base system Signed-off-by: Maximilian Meister
* Sat Jan 27 2018 containers-bugownerAATTsuse.de- Commit 46f651b by Rafael Fernández López ereslibreAATTereslibre.es Adapt documentation Commit b4314de by Rafael Fernández López ereslibreAATTereslibre.es Fix specs Commit d91b79f by Rafael Fernández López ereslibreAATTereslibre.es Update velum spec and remove assets group Commit 534605e by Rafael Fernández López ereslibreAATTereslibre.es Only require the relevant gem groups Commit 8dd2061 by Rafael Fernández López ereslibreAATTereslibre.es Remove `kubeconfig`. It\'s an unused dependency. Commit 103447f by Rafael Fernández López ereslibreAATTereslibre.es Add assets as another group Commit da7e999 by Rafael Fernández López ereslibreAATTereslibre.es Remove PACKAGING envvar Commit a723afc by Maximilian Meister mmeisterAATTsuse.de use bundler to extract build requires Signed-off-by: Maximilian Meister Commit 89c90f3 by Rafael Fernández López ereslibreAATTereslibre.es Fix packaging by forcing the subset of gems to be installed. Our `Gemfile` and `Gemfile.lock` should be the same under all environments, development, test and production. Using an envvar on the Gemfile to decide the dependency tree might cause deep problems and differences on the dependency tree between development, test and production. We want to have the same gem dependency tree on all possible environments, so we don\'t hit bugs in production caused by a different dependency tree than the one used in development and test, as this bugs could only be hit in production and never reproduced in development and test environments.
* Tue Jan 23 2018 containers-bugownerAATTsuse.de- Commit b6920e1 by Maximilian Meister mmeisterAATTsuse.de drop workaround for old bundler versions in caasp we are using the new bundler as well now Signed-off-by: Maximilian Meister
* Mon Jan 22 2018 containers-bugownerAATTsuse.de- Commit d4b52de by Maximilian Meister mmeisterAATTsuse.de enhance specfile Signed-off-by: Maximilian Meister
* Mon Jan 22 2018 containers-bugownerAATTsuse.de- Commit e13d260 by Rafael Fernández López ereslibreAATTereslibre.es Do not require `retcode` to be 0 to detect success. When detecting highstate success as well as orchestration success we are checking the result success status and also checking that the global `retcode` is 0. This latest check is not strictly necessary and under certain conditions it is not desirable.
* Fri Jan 19 2018 containers-bugownerAATTsuse.de- Commit 5bcf8be by Maximilian Meister mmeisterAATTsuse.de adapt devel libraries to tumbleweed in tumbleweed it\'s named mariadb, in sle it\'s mysql Signed-off-by: Maximilian Meister
* Fri Jan 19 2018 containers-bugownerAATTsuse.de- Commit 14d4c91 by Maximilian Meister mmeisterAATTsuse.de re-add the patch to relax a gem requirement rubygem-kubeclient upstream meanwhile updated their requirements as well Signed-off-by: Maximilian Meister
* Fri Jan 19 2018 containers-bugownerAATTsuse.de- Commit d6f9c74 by Rafael Fernández López ereslibreAATTereslibre.es Add migration to change the pillar value column to text
* Thu Jan 18 2018 containers-bugownerAATTsuse.de- Commit 4994345 by James Mason jmasonAATTsuse.com Consistent representation of public cloud instance types In order to present a reasonable set of choices for instance_type, a collection of instance types was hand-picked for each supported public cloud framework.
* Thu Jan 18 2018 containers-bugownerAATTsuse.de- Commit 8bab191 by Maximilian Meister mmeisterAATTsuse.de bump gems for rails 4.2.10 Signed-off-by: Maximilian Meister
* Thu Jan 18 2018 containers-bugownerAATTsuse.de- Commit 3a01616 by Maximilian Meister mmeisterAATTsuse.de Update rails to 4.2.10
* Wed Jan 17 2018 containers-bugownerAATTsuse.de- Commit a3d0e50 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de fix test for 2.5 in travis by updating the rubygems see https://github.com/travis-ci/travis-ci/issues/8978
* Tue Jan 16 2018 containers-bugownerAATTsuse.de- Commit 2aba80f by Rafael Fernández López ereslibreAATTereslibre.es Allow to retry the upgrade orchestration
* Tue Jan 16 2018 containers-bugownerAATTsuse.de- Commit f4f6c24 by Rafael Fernández López ereslibreAATTereslibre.es Serve registry mirrors as salt expects it
* Mon Jan 15 2018 containers-bugownerAATTsuse.de- Commit 14f8ad8 by Vítor Avelino contactAATTvitoravelino.me Added SUSE registry mirror settings to bootstrap page With this patch we allow our users to configure mirrors poiting to to the SUSE registry. It can handle insecure and secure mirrors with optional certificate.
* Mon Jan 15 2018 containers-bugownerAATTsuse.de- Commit efa6d18 by Rafael Fernández López ereslibreAATTereslibre.es Fix version for development Fixes: bsc#1071475
* Fri Jan 12 2018 containers-bugownerAATTsuse.de- Commit cc9b32a by Maximilian Meister mmeisterAATTsuse.de packaging: enhancements
* drop removal of .orig files we should keep patches rebased instead
* rebase patches
* drop empty scriptlets this was feedback from factory submission
* reduce verbosity this massively reduces the build log
* Thu Jan 11 2018 containers-bugownerAATTsuse.de- Commit f491b2f by Rafael Fernández López ereslibreAATTereslibre.es Add `CertificateService` proxy model that will tie `Certificate` to any service
* Wed Jan 10 2018 containers-bugownerAATTsuse.de- Commit f39b7a2 by Rafael Fernández López ereslibreAATTereslibre.es Set default flannel backend to UDP for old installations. This migration will only be executed on old installations, while on new ones (where we will use the default salt pillar value) this migration won\'t be run, so there won\'t be a pillar override.
* Wed Jan 10 2018 containers-bugownerAATTsuse.de- Commit 20fedda by Flavio Castelli fcastelliAATTsuse.com AutoYaST: disable creation of swap [bsc#1043813] Disables the creation of the swap partition on the nodes provisioned by AutoYaST. This fixes bsc#1075001 Signed-off-by: Flavio Castelli
* Fri Dec 22 2017 containers-bugownerAATTsuse.de- Commit b43e58d by James Mason jmasonAATTsuse.com Add a rake task for inserting a pillar Due to redundancy, argument validation of velum:create_user was abstracted out to a separate function. Commit 8781733 by James Mason jmasonAATTsuse.com Add csp pillar
* Wed Dec 20 2017 containers-bugownerAATTsuse.de- Commit c2aabd4 by Rafael Fernández López ereslibreAATTereslibre.es Reject `nil` values on the Velum pillar. This gives a chance for default salt pillars to be used.
* Tue Dec 19 2017 containers-bugownerAATTsuse.de- Commit 488e08c by Kiall Mac Innes kiallAATTmacinnes.ie Remove dev vs prod kubeconfig download difference This was needed to workaround capybara + poltergeist being unable to download files, this workarond is now down in the velum-interactions tooling, rather than in a shiped part of the codebase.
* Tue Dec 05 2017 containers-bugownerAATTsuse.de- Commit cea27bc by Maximilian Meister mmeisterAATTsuse.de rebase patch Signed-off-by: Maximilian Meister
* Fri Dec 01 2017 containers-bugownerAATTsuse.de- Commit 81cf5c1 by Maximilian Meister mmeisterAATTsuse.de rebase patch again since rake changed again Signed-off-by: Maximilian Meister
* Fri Dec 01 2017 containers-bugownerAATTsuse.de- Commit 4a14086 by Maximilian Meister mmeisterAATTsuse.de rebase patch Signed-off-by: Maximilian Meister
* Wed Nov 29 2017 containers-bugownerAATTsuse.de- Commit fa1d18c by Rafael Fernández López ereslibreAATTereslibre.es Remove unused factory
* Tue Nov 28 2017 containers-bugownerAATTsuse.de- Commit 9aca2ce by Rafael Fernández López ereslibreAATTereslibre.es Allow to restart orchestrations This commit includes a few changes:
* `Orchestration` model has been introduced
* We have a history of all the orchestrations that happened, what type of orchestration it was, how long it took to be completed and the result of the orchestration.
* `Retryable?` concept has been included in `Orchestration`
* Right now, only the bootstrap retryable concept has been included in the UI, but the backend supports the update orchestration too.
* `SaltOrchestration` proxy class has been included, so it can retrieve some information from salt-api directly (accessed through the Orchestration model).
* Highstates now don\'t mark minions as successful. Highstates now can only mark minions as failed if the highstate failed.
* Only the orchestrations will now mark success when it finishes. All pending minions will be marked as succeeded when this happens.
* External orchestrations are detected (we can trigger them completely outside Velum and it will detect there is an orchestration running). This is because of the introduced salt handler.
* `MinionOrchestration` handler has been renamed to `OrchestrationResult`.
* Helper methods have been introduced to `Velum::Salt`, so it\'s easier to retrieve the list of jobs as well as the detail of one job, given its `jid`.
* `salt-api` http calls has now always the same options as in `development` and `production` when executed in a `test` environment. Slowness problems were because of flaky tests with Capybara waiting too much on things we weren\'t expecting.
* Some VCR cassettes and tests have been added. Fixes: bsc#1069154
* Tue Nov 28 2017 containers-bugownerAATTsuse.de- Commit a6a3159 by Rafael Fernández López ereslibreAATTereslibre.es Remove unneeded unprotected pillars There\'s no possible use case for these pillars in which they were filled with information and then the customer will clear them (as they are required). Keep things as simple as possible until we refactor this part. Without this change, a specially crafted HTTP request could be made to remove the content of those pillar values, when this is not ideal.
* Tue Nov 28 2017 containers-bugownerAATTsuse.de- Commit f280bc6 by Dario Maiocchi dmaiocchiAATTsuse.com Omit the default :each argument for RSpec hooks.
* Tue Nov 28 2017 containers-bugownerAATTsuse.de- Commit 4a17aa7 by Vítor Avelino contactAATTvitoravelino.me js: fixed alert color being overridden The alert notifying the user that the bootstrap has failed was overriding the type of any previous alert that were on the page. This was causing some confusing to some users. Now it creates an alert everytime we check if there\'s a failed bootstrapped node without interfering other alerts. Also removed unnecessary stuff from the notification partial. Fixes bsc#1069258
* Mon Nov 27 2017 containers-bugownerAATTsuse.de- Commit c5637d2 by Rafael Fernández López ereslibreAATTereslibre.es Do not expose puma directly
* Thu Nov 23 2017 containers-bugownerAATTsuse.de- Commit c753ed6 by Maximilian Meister mmeisterAATTsuse.de update puma to 3.10.0 this includes a fix for lockup off ssl connections Signed-off-by: Maximilian Meister
* Tue Nov 21 2017 containers-bugownerAATTsuse.de- Commit fdb9fcb by Rafael Fernández López ereslibreAATTereslibre.es Implement Pillar endpoint and introduce internal-api Fixes: bsc#1069145
* Tue Nov 07 2017 containers-bugownerAATTsuse.de- Commit a3b3784 by Rafael Fernández López ereslibreAATTereslibre.es When we are back in Velum from Dex, allow for assets to be loaded If we forced the location directly, javascript assets were never loaded as their loading are interrupted by the browser because of the `href` change. This ensures we first load all the assets (so delete, put, patch links work through jquery as usual), and then we force the redirection to happen. Fixes: bsc#1066611
* Mon Nov 06 2017 containers-bugownerAATTsuse.de- Commit c54fbc9 by Rafael Fernández López ereslibreAATTereslibre.es Only allow `development` environment to override the database with `VELUM_DB_ENV`. Test and production database names are fixed.
* Mon Nov 06 2017 containers-bugownerAATTsuse.de- Commit b07fde3 by Vítor Avelino contactAATTvitoravelino.me ux: changed color of pending status icon Changed from green to gray. The reason behind this is that green is usually associated with a success operation and gray is a more neutral color that doesn\'t add much meaning.
* Mon Nov 06 2017 containers-bugownerAATTsuse.de- Commit dc92f64 by Vítor Avelino contactAATTvitoravelino.me js: fixed undesired minimum nodes alert blink Moved the calls of `toggleMinimumNodesAlert` and `handleBootstrapErrors` to be done after the checkbox changes to `true` instead of by any change (false/true). With this change there will be no middle state that cause the alert when `toggleMinimumNodesAlert` is called. Fixes bsc#1066371
* Fri Nov 03 2017 containers-bugownerAATTsuse.de- Commit e7d877f by Maximilian Meister mmeisterAATTsuse.de handle invalid sessions more user friendly if the session is not valid anymore, show this information to the user, and redirect him to the login instead of failing badly we dont need to raise in the stored_nonce method, as a failed deletion from the session hash will result in a nil value, meaning a non matching nonce Signed-off-by: Maximilian Meister
* Tue Oct 31 2017 containers-bugownerAATTsuse.de- Commit dbbaff5 by Rafael Fernández López ereslibreAATTereslibre.es Add `velum:migrate_users` task that will migrate database-backed users into LDAP-backed users. In 1.0 user login information was stored in the database whereas in 2.0 this information resides in LDAP. Fixes: bsc#1065450
* Tue Oct 31 2017 containers-bugownerAATTsuse.de- Commit 15d26a7 by Vítor Avelino contactAATTvitoravelino.me Fixed role column broken up on two lines Fixes bsc#1064807
* Mon Oct 30 2017 containers-bugownerAATTsuse.de- Commit 3ee3aa4 by Rafael Fernández López ereslibreAATTereslibre.es Add kubeconfig tips on the download page. Also, add a link to the `root_path` on the icon of CaaS Platform. Fixes: bsc#1062728
* Mon Oct 30 2017 containers-bugownerAATTsuse.de- Commit adc2cb4 by Maximilian Meister mmeisterAATTsuse.de enable remembering users fixes bsc#1062184 Signed-off-by: Maximilian Meister
* Mon Oct 30 2017 containers-bugownerAATTsuse.de- Commit 73c5544 by Rafael Fernández López ereslibreAATTereslibre.es Make Dex first redirect to Velum, and download the kubeconfig file from there Fixes: bsc#1062728
* Fri Oct 27 2017 containers-bugownerAATTsuse.de- Commit 758d9f5 by Vítor Avelino contactAATTvitoravelino.me Improved failure alert message This is an attempt to make the experience better. We thought that the previous message was too aggressive. Since this message will always be shown for 2.0 at least, we want to tranquilize the user that it\'s an expected behavior.
* Fri Oct 27 2017 containers-bugownerAATTsuse.de- Commit a30a9df by Vítor Avelino contactAATTvitoravelino.me typo: interally -> internally Small typo on \"Overlay network settings\" settings. Fixes bsc#1065265
* Thu Oct 26 2017 containers-bugownerAATTsuse.de- Commit 3151ad8 by Maximilian Meister mmeisterAATTsuse.de alert the user when bootstrap failed otherwise the user is left without any clue what to do Signed-off-by: Maximilian Meister
* Tue Oct 24 2017 containers-bugownerAATTsuse.de- Commit 574b502 by Rafael Fernández López ereslibreAATTereslibre.es When roles are set, set `not_applied` if it\'s a local assignment. If we are setting the role locally only (on the database), set the minion status to `not_applied`. If we are also setting the remote role (set on salt too), set the minion status to `pending`. Since there\'s no Orchestration model yet, there\'s no entity to control this assignments, and Minion model will set one or the other depending on the `remote` attribute of `assign_role`. Fixes: bsc#1063977
* Tue Oct 24 2017 containers-bugownerAATTsuse.de- Commit 2cba6d1 by Vítor Avelino contactAATTvitoravelino.me Notify user when not able to fetch data from Velum API We are regenerating Velum\'s certificate by adding the external dashboard fqdn/ip to the Velum certificate. For this certificate to be reloaded we are currently doing a `docker restart` on Velum while the orchestration is being executed. This has a bad effect on the user session as he/she is waiting for the polling to finish. Without them noticing, the polling will start failing as the certificate changed under them, and they need to accept it, but ajax requests will just fail because of this. With this patch we are notifying the user that the browser cannot connect with Velum API and suggests he/she to try to reload the session. Fixes bsc#1064641
* Mon Oct 23 2017 containers-bugownerAATTsuse.de- Commit 450e6bc by Rafael Fernández López ereslibreAATTereslibre.es Reuse filter logic for known hosts. Also, do not repeat the allowed hosts if they match.
* Fri Oct 20 2017 containers-bugownerAATTsuse.de- Commit 5255b9a by Maximilian Meister mmeisterAATTsuse.de spec: make coverage folder public this has the big advantage that you can check the results directly in velum under http(s)://$velum/coverage/index.html since we switched the official devenv to containers within VM\'s the folder is not accessible that easy anymore Signed-off-by: Maximilian Meister
* Thu Oct 19 2017 containers-bugownerAATTsuse.de- Commit 87be131 by Rafael Fernández López ereslibreAATTereslibre.es When an orchestration result is received, update failed nodes too. Consider all pending and failed minions when we receive a kubernetes orchestration result. If later orchestrations succeed they should also mark failed nodes as succeeded. Fixes: bsc#1064281
* Thu Oct 19 2017 containers-bugownerAATTsuse.de- Commit 5401e86 by Maximilian Meister mmeisterAATTsuse.de make sure request comes from a registered host warn when user accesses from unknown fqdn, and log out when the kubectl download is requested via unknown fqdn bsc#1062728 Signed-off-by: Maximilian Meister
* Thu Oct 19 2017 containers-bugownerAATTsuse.de- Commit 835aa11 by Maximilian Meister mmeisterAATTsuse.de better explain internal dashboard fqdn bsc#1063233 Signed-off-by: Maximilian Meister
* Tue Oct 17 2017 containers-bugownerAATTsuse.de- Commit 76a71f4 by Kiall Mac Innes kiallAATTmacinnes.ie Revert \"Set etcd2 backend for old installations\" This reverts commit 25a0658f22db1f818403f01cc8bc97a36c772fad.
* Tue Oct 17 2017 containers-bugownerAATTsuse.de- Commit a77184d by Maximilian Meister mmeisterAATTsuse.de Correctly display the logo SVG had wrong width, 300 instead of 314. For some reason the text appeared cropped only in Firefox. After saving the SVG with the correct width, logo now is displayed properly. bsc#1062832 Signed-off-by: Maximilian Meister converted svg font object to path
* Fri Oct 13 2017 containers-bugownerAATTsuse.de- Commit 73e86cb by Maximilian Meister mmeisterAATTsuse.de fix rubocop offense Signed-off-by: Maximilian Meister Commit ed411fd by Maximilian Meister mmeisterAATTsuse.de fill in some values and wait for button to appear Signed-off-by: Maximilian Meister
* Thu Oct 12 2017 containers-bugownerAATTsuse.de- Commit d871bac by Maximilian Meister mmeisterAATTsuse.de also compile svg the logo has recently changed to svg and needs to be compiled too otherwise it wont be shown bsc#1062878 Signed-off-by: Maximilian Meister
* Wed Oct 11 2017 containers-bugownerAATTsuse.de- Commit c03cd33 by Kiall Mac Innes kiallAATTmacinnes.ie Collect the dashboard hosts external FQDN This should be used by Dex to redirect back to Velum, allowing for scenarios where the admin node\'s private IP is not accessible to the outside world - e.g. as is the case on OpenStack and possibly other cloud environments. Part of bsc#1062291
* Tue Oct 10 2017 containers-bugownerAATTsuse.de- Commit c1f1f1e by Rafael Fernández López ereslibreAATTereslibre.es Wait for orchestration result for marking success on minions We are marking success depending on highstates, what makes a bit misleading for the overall orchestration process. At this moment, orchestration will wait for some tasks to have been completed (like some `kubectl apply` commands, which apply immediately). We aren\'t waiting for the global orchestration result, and this means that all minions might be marked as successful when we haven\'t really waited for the overall orchestration to have finished. This can have important implications, like not waiting for Dex to be completely ready while the orchestration is waiting for it. We just mark failed highstates as an early failure, but wait for the orchestration result to finish and be succesful in order to mark pending nodes as successful. This will also align with the current restritions in the UI to enable the `kubectl config` button, since it will wait for the kubernetes master to be ready in order to be enabled. Fixes: bsc#1062542
* Tue Oct 10 2017 containers-bugownerAATTsuse.de- Commit b2ed55d by Maximilian Meister mmeisterAATTsuse.de add new svg logo with the full name SUSE® CaaS Platform bsc#1062247 Signed-off-by: Maximilian Meister
* Sat Oct 07 2017 containers-bugownerAATTsuse.de- Commit 5c0a6c7 by Kiall Mac Innes kiallAATTmacinnes.ie Update VERSION file to 2.0.0+dev
* Fri Oct 06 2017 containers-bugownerAATTsuse.de- Commit c77b369 by Rafael Fernández López ereslibreAATTereslibre.es Allow to serve the kubeconfig file in a raw fashion if the environment is not production Commit b6ba0d3 by Robert Roland robert.rolandAATTsuse.com Allowing a user to download a full kubeconfig Works by having the user perform an OIDC handshake against the running cluster, then generating a kubeconfig based on that handshake bsc#1058568
* Thu Oct 05 2017 containers-bugownerAATTsuse.de- Commit 1df8e6e by Vítor Avelino contactAATTvitoravelino.me ux: improved messages for \'Next\' button on nodes selection The title message for the \'Next\' button on the nodes selection page were out of date and weren\'t helping the user understand what was wrong. For each rule we use to validate the \'Next\' button we are adding a dedicated message to show to the user. This way the user will know exactly what\'s wrong with the cluster configuration. I also fixed the \"minimum of three nodes\" alert to only be displayed if the user didn\'t select 3 nodes (dones\'t matter if master was not selected). The previous behavior was making it display if the configuration wasn\'t supported even having 4 nodes selected. This info now goes on the \'Next\' button title message. See bsc#1058915
* Wed Oct 04 2017 containers-bugownerAATTsuse.de- Commit 52abdd5 by Rafael Fernández López ereslibreAATTereslibre.es
* Fix syntax error
* Move helm tiller option to the first page of the setup
* Recover the set value if the page is reloaded
* Tue Sep 19 2017 containers-bugownerAATTsuse.de- Commit bd44274 by Rafael Fernández López ereslibreAATTereslibre.es Do not add an override that will have the same value as the overriden one This will only be effective on new installations, where the default pillar value matches this value. Let\'s keep things simple and only override what is in the UI or what will/can be different between versions.
* Mon Sep 18 2017 containers-bugownerAATTsuse.de- Commit 0cf2d39 by Rafael Fernández López ereslibreAATTereslibre.es Allow to provision cluster nodes with autoyast in an air-gapped environment Fixes: bsc#1058999
* Mon Sep 18 2017 containers-bugownerAATTsuse.de- Commit e6d76c by Rafael Fernández López ereslibreAATTereslibre.es Revert \"Serve the autoyast content even when a SCC/SMT connection err… …or happens\" This reverts commit 1bd1de5
* Fri Sep 15 2017 containers-bugownerAATTsuse.de- Commit 7625b36 by Rafael Fernández López ereslibreAATTereslibre.es Consistent naming of the product - SUSE CaaSP - SUSE Container as a Service Platform both should be SUSE CaaS Platform. Fixes: bsc#1058720
* Fri Sep 15 2017 containers-bugownerAATTsuse.de- Commit d2aa99d by Kiall Mac Innes kiallAATTmacinnes.ie Revert \"Adding option to install Tiller\" This reverts commit f329dd7969f1ad7635b9e297b838e447fff3faa8.
* Wed Sep 13 2017 containers-bugownerAATTsuse.de- Commit 0e9d610 by Robert Roland robert.rolandAATTsuse.com LDAP - remove hardcoded bind password Passes the LDAP parameters into the pillar so Dex can be bootstrapped properly
* Tue Sep 12 2017 containers-bugownerAATTsuse.de- Commit 1bd1de5 by Rafael Fernández López ereslibreAATTereslibre.es Serve the autoyast content even when a SCC/SMT connection error happens The service unavailable http error code (503) will only be returned now until the first page of the setup is completed (so we have the dashboard address to serve the autoyast profile). When one of the following exceptions happens: - Missing registration code - Missing credentials for SCC/SMT - Any HTTP error while connecting to SCC/SMT We will still serve the autoyast content with a comment explaining why the system cannot be registered. The returned http error code will be success (200), but we will show a dialog stating the problem that was detected and why it won\'t be registered. Fixes: bsc#1049948
* Tue Sep 12 2017 containers-bugownerAATTsuse.de- Commit ed94e9a by Vítor Avelino contactAATTvitoravelino.me Fixed proxy settings when revisiting the page After the first step of setting up the proxy settings, when the user goes back to that initial page, the proxy settings were lost. This can not only confuse the user but also introduce undesired changes. With this fix it doesn\'t happen anymore. Fixes #300
* Tue Sep 12 2017 containers-bugownerAATTsuse.de- Commit c084ca4 by Rafael Fernández López ereslibreAATTereslibre.es Only override blank pillars on the page they are shown Before this change the last step of the setup would remove the pillar information for the first step proxy values (since they were not protected pillars). Instead, explicitly say what unprotected pillars there are in each step and allow the controller to override them with blank if they were provided blank after having some other value.
* Tue Sep 12 2017 containers-bugownerAATTsuse.de- Commit c73cd4e by Rafael Fernández López ereslibreAATTereslibre.es Add networking settings to the setup UI
* Add to the pillar the old settings. This migration will only be executed on existing installations (as the `schema.rb` defines the up to date version and this last migration will be excluded)
* Tue Sep 12 2017 containers-bugownerAATTsuse.de- Commit 8c73bdd by Rafael Fernández López ereslibreAATTereslibre.es Remove explicit set for `pidfile` and `statefile`
* Tue Sep 12 2017 containers-bugownerAATTsuse.de- Commit 25a0658 by Rafael Fernández López ereslibreAATTereslibre.es Set etcd2 backend for old installations
* Fri Sep 08 2017 containers-bugownerAATTsuse.de- Commit 0f8ef1f by Vítor Avelino contactAATTvitoravelino.me Fixed \'Accept All Nodes\' button ajax request The \'Accept All Nodes\' button had the class \'.accept-minion\' that is used to accept individually a node. When clicking on it the wrong event listener callback was being called instead of the correct one. Now it calls the expected event listener callback and makes the request passing the expected arguments for all the nodes to be accepted.
* Fri Sep 08 2017 containers-bugownerAATTsuse.de- Commit f329dd7 by Sam Leavens rbwsamAATTgmail.com Adding option to install Tiller
* Wed Sep 06 2017 containers-bugownerAATTsuse.de- Commit 89165f5 by Robert Roland rob.rolandAATTgmail.com RBAC: Adding LDAP support to Velum (#286)
* Adding LDAP support to Velum
* Will create the requisite org units in LDAP if they are missing
* Will create users in LDAP if they are created via Velum registration
* Tue Sep 05 2017 containers-bugownerAATTsuse.de- Commit b1c911f by Rafael Fernández López ereslibreAATTereslibre.es Clean old jobs on our own, since salt won\'t do it automatically Given our current configuration in which we are no longer using the mysql `job_cache`, salt won\'t clean old jobs on its own, so we should do it so the database won\'t keep growing without control. We are executing raw SQL commands to avoid creating `Jid` and `SaltReturn` models just for the purpose of cleaning, as we know that we want to move away from a mysql backed salt master. This is a follow up of: https://github.com/kubic-project/salt/pull/199 Fixes: bsc#1054256
* Tue Sep 05 2017 containers-bugownerAATTsuse.de- Commit 5e20ff5 by Maximilian Meister mmeisterAATTsuse.de fix travis for mysql-devel Signed-off-by: Maximilian Meister
* Tue Sep 05 2017 containers-bugownerAATTsuse.de- Commit ef0160f by Maximilian Meister mmeisterAATTsuse.de packaging: mention velum-rpmlintrc in the sources openSUSE Factory review bot is declining requests because of that Signed-off-by: Maximilian Meister
* Tue Sep 05 2017 containers-bugownerAATTsuse.de- Commit 2de7e3a by Kiall Mac Innes kiallAATTmacinnes.ie Add a 200MB EFI partition to AutoYaST XML Fixes bsc#1055795
* Mon Sep 04 2017 containers-bugownerAATTsuse.de- Commit 6ce1dda by Rafael Fernández López ereslibreAATTereslibre.es Add rake task velum:create_user This rake task takes two parameters: an email and a password, and will create a user entry in the Velum database so it\'s possible to login and work with it.
* Mon Aug 28 2017 containers-bugownerAATTsuse.de- Commit dc79e32 by Rafael Fernández López ereslibreAATTereslibre.es Modify Velum workflow for Kubernetes external FQDN name Velum will now have a new setup step after the roles have been chosen, with the selected master as the kubernetes external FQDN name that will contain its FQDN by default, allowing the customer to modify it to fit their needs. Fixes: bsc#1047284
* Thu Aug 17 2017 containers-bugownerAATTsuse.de- Commit 77c6d77 by Vítor Avelino contactAATTvitoravelino.me removed es6 syntax; fixed tests; added new tests Commit 5295c14 by Vítor Avelino contactAATTvitoravelino.me replaced checkboxes by button group Commit 2151cd1 by Kiall Mac Innes kiallAATTmacinnes.ie Allow multiple master selection Allow the user to choose multiple masters during the initial bootstrap. We don\'t currently allow for adding a master post bootstrap.
* Wed Aug 09 2017 containers-bugownerAATTsuse.de- Commit caef239 by Flavio Castelli fcastelliAATTsuse.com Set database name via environment variable Allow to specify a custom database name using the `VELUM_DB_NAME` environment variable. The code falls back to the old value when the environment variable is not set. Signed-off-by: Flavio Castelli
* Fri Aug 04 2017 containers-bugownerAATTsuse.de- Commit b193e5e by Robert Roland robert.rolandAATTsuse.com Small grammar update I know it\'s nitpicky, but my OCD requires me to finally change this text.
* Thu Aug 03 2017 containers-bugownerAATTsuse.de- Commit 4d07cc5 by Maximilian Meister mmeisterAATTsuse.de allow kubeconfig download when master is ready (bsc#1048483) Signed-off-by: Maximilian Meister
* Tue Aug 01 2017 containers-bugownerAATTsuse.de- Commit 0d08127 by Maximilian Meister mmeisterAATTsuse.de reduce 503 gap (bsc#1049931) there is a gap between the database setup and the start of the rails server which can be reduced Signed-off-by: Maximilian Meister
* Mon Jul 31 2017 containers-bugownerAATTsuse.de- Commit 2386314 by Maximilian Meister mmeisterAATTsuse.de set secure flag in cookie (bsc#1050484) Signed-off-by: Maximilian Meister
* Tue Jul 25 2017 containers-bugownerAATTsuse.de- Commit 3f44180 by Christoph Wickert cwickertAATTsuse.de Fix loopback IP for proxy exception during initial configuration (127.0.01 -> 127.0.0.1)
* Fri Jul 21 2017 containers-bugownerAATTsuse.de- Commit 7a49514 by Vítor Avelino contactAATTvitoravelino.me bsc#1049165 : avoid orchestration action from ui while updating/adding node - When an admin node update is available, do not allow users to press \"update all nodes\". The admin node should be updated first. - Never allow both a workers update and an admin update to be started at the same time. - Velum should not allow add node while a bootstrap/upgrade/etc is in progress
* Fri Jul 21 2017 containers-bugownerAATTsuse.de- Commit 9862097 by Vítor Avelino contactAATTvitoravelino.me bsc#1045734: fix show/hide update link based on update/highstate `Update all nodes` link should be shown if all nodes are in applied state and there\'s an update available/retryable. If for some reason a node goes to a different state than \"applied\" or there\'s no update, link becomes unavailable.
* Tue Jul 18 2017 containers-bugownerAATTsuse.de- Commit 9308c52 by Maximilian Meister mmeisterAATTsuse.de increase workers (bsc#1048533) this allows for more concurrent api requests a worker spawns a copy of the velum process Signed-off-by: Maximilian Meister
* Mon Jul 17 2017 containers-bugownerAATTsuse.de- Commit 9448526 by Kiall Mac Innes kiallAATTmacinnes.ie Handle failues of the update orchestration Handle the failure of an update orchestration in the same way we handle a failure for a bootstrap orchestration. bsc#1047895
* Fri Jul 14 2017 containers-bugownerAATTsuse.de- Commit a285f00 by Maximilian Meister mmeisterAATTsuse.de make branch safe by transforming slashes to dashes Signed-off-by: Maximilian Meister (cherry picked from commit 49ab27251e452026bee834eb56709180c781a783) Commit 7e861b6 by Maximilian Meister mmeisterAATTsuse.de packaging: make branch configurable Signed-off-by: Maximilian Meister (cherry picked from commit b0d65b5acc4beea1052e748120fbbf42a8715fc4)
* Fri Jul 14 2017 containers-bugownerAATTsuse.de- Commit aed8918 by Rafael Fernández López ereslibreAATTereslibre.es Detect orchestration results when `mods` has been provided as argument We usually run the orchestrations manually as: `salt-run state.orchestrate orch.kubernetes` And this case was detected properly. This change adds support for also detecting the orchestration result when the orchestration is called using the `salt-api` on Velum (the general case): `salt-run state.orchestrate mods=orch.kubernetes` With this change we support both return results. Fixes: bsc#1046706
* Thu Jul 13 2017 containers-bugownerAATTsuse.de- Commit aed8918 by Rafael Fernández López ereslibreAATTereslibre.es Detect orchestration results when `mods` has been provided as argument We usually run the orchestrations manually as: `salt-run state.orchestrate orch.kubernetes` And this case was detected properly. This change adds support for also detecting the orchestration result when the orchestration is called using the `salt-api` on Velum (the general case): `salt-run state.orchestrate mods=orch.kubernetes` With this change we support both return results. Fixes: bsc#1046706
* Wed Jul 12 2017 containers-bugownerAATTsuse.de- Commit 4ae4648 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de [packaging] rebase schema patch because of the changes introduced in 7e2e454 Signed-off-by: Jordi Massaguer Pla
* Wed Jul 12 2017 containers-bugownerAATTsuse.de- Commit 79cccb5 by Rafael Fernández López ereslibreAATTereslibre.es `force_ssl` is documented to be executed from the class context Not all redirections to `https` were properly handled. By making this change we manually redirect to `https` as a `before_action` as before, but handling a `redirect_to` in the chain (effectively stopping the chain and in the expected order). The reason behind this is that `force_ssl` called from our own filter: https://github.com/rails/rails/blob/03925dc26a747075ff37660d1f0a060a8178bf66/actionpack/lib/action_controller/metal/force_ssl.rb#L62-L68 installs a `before_action`, but it\'s already too late for that very request (we are already in the `before_actions` for that request). Also, with the current implementation the filter gets installed after our other `before_action` filters, what implies that the redirection to `https` is not the very first filter in the chain, as it should be. By calling to the method itself: https://github.com/rails/rails/blob/03925dc26a747075ff37660d1f0a060a8178bf66/actionpack/lib/action_controller/metal/force_ssl.rb#L76-L96 from our filter, we ensure all requests are redirected to TLS. Fixes: bsc#1048134
* Wed Jul 12 2017 containers-bugownerAATTsuse.de- Commit 55895e9 by Rafael Fernández López ereslibreAATTereslibre.es Add missing indexes for `salt_events` table Fix some slow queries that we were having due to missing indexes. Increase batch job size. Fixes: bsc#1046862
* Tue Jul 11 2017 containers-bugownerAATTsuse.de- Commit a40ed8a by Kiall Mac Innes kiallAATTmacinnes.ie Add Jenkinsfile The Jenkinsfile in each repo, if we adopt Jenkins in the end, will be very thin, including just a single library load, and a single method call. This prevents us from needing to keep each projects Jenkinsfile in sync as CI changes are made.
* Fri Jul 07 2017 containers-bugownerAATTsuse.de- Commit c28c294 by Flavio Castelli fcastelliAATTsuse.com db not ready page: return http 503 The \"db is not ready\" page now returns a http 503 status instead of a 200. Signed-off-by: Flavio Castelli
* Fri Jul 07 2017 containers-bugownerAATTsuse.de- Commit 054e21c by Rafael Fernández López ereslibreAATTereslibre.es Take orchestration result into consideration when bootstrapping When bootstrapping, if the orchestration fails, depending in what stage it failed, we might never retrieve individual highstate returns from minions, so they will be in `pending` status forever. With this change, we also listen for the overall bootstrap orchestration result (`orch.kubernetes` for now), so we will mark all pending nodes as failed. Fixes: bsc#1046706
* Fri Jul 07 2017 containers-bugownerAATTsuse.de- Commit 53f71da by Rafael Fernández López ereslibreAATTereslibre.es Before checking all nodes, make sure we are waiting for them When using the `check-all` box, make sure that all expected nodes are present on the page, so we select all the nodes that we intend to, instead of only a subset of them. This should help remove flaky tests. Commit bab4226 by Miquel Sabaté Solà msabateAATTsuse.com Improved the reliability of some tests There\'s a flaky test that has been marked as `pending` since: 1. It\'s already tested as an end-to-end test. 2. We\'ve wasted too much time on it already and we have more pressing issues. Besides that, we\'ve applied best practices and tests should run faster overall. This commit has been done with the help of AATTvitoravelino and AATTereslibre. Signed-off-by: Miquel Sabaté Solà
* Fri Jul 07 2017 containers-bugownerAATTsuse.de- Commit f34180e by Flavio Castelli fcastelliAATTsuse.com Serve \"db is not ready\" page over https [bsc#1047203] Ensure the \"db is not yet ready\" is served over https too. This fixes bsc#1047203. Signed-off-by: Flavio Castelli
* Fri Jul 07 2017 containers-bugownerAATTsuse.de- Commit 7f8256d by Maximilian Meister mmeisterAATTsuse.de change dark to light font color in the dark background Signed-off-by: Maximilian Meister Commit 68927b8 by Maximilian Meister mmeisterAATTsuse.de public: style the other static pages as well follow up of #237 Signed-off-by: Maximilian Meister
* Thu Jul 06 2017 containers-bugownerAATTsuse.de- Commit 67308fb by Flavio Castelli fcastelliAATTsuse.com Improve the style of \"db is initializing\" page [bsc#1047305] Try to use the same style as the main velum web application. Signed-off-by: Flavio Castelli
* Thu Jul 06 2017 containers-bugownerAATTsuse.de- Commit b898f75 by Rafael Fernández López ereslibreAATTereslibre.es Download the kubectl client certificate and key from the new location Fixes: bsc#1046963
* Thu Jul 06 2017 containers-bugownerAATTsuse.de- Commit 2211633 by PI-Victor palade.ionutAATTgmail.com views: remove lore ipsum dead code from worker_bootstrap.html.slim
* Wed Jul 05 2017 containers-bugownerAATTsuse.de- Commit 125fb16 by Rafael Fernández López ereslibreAATTereslibre.es Make travis load the required phantomjs version Also improve caching for faster builds.
* Wed Jul 05 2017 containers-bugownerAATTsuse.de- Commit 20c9e03 by Rafael Fernández López ereslibreAATTereslibre.es Fix versioning syntax for ruby patchset version
* Wed Jul 05 2017 containers-bugownerAATTsuse.de- Commit 74b395e by Rafael Fernández López ereslibreAATTereslibre.es Add two runs for each ruby version, one with fail fast, the other a regular one When we find a flaky test, the fail fast version will contain relevant information on the tail of the `test.log` fail, while the regular run will continue to show any other errors. Also, use the ruby version used in CaaSP only for now.
* Wed Jul 05 2017 containers-bugownerAATTsuse.de- Commit 557c490 by Rafael Fernández López ereslibreAATTereslibre.es Print `log/test.log` upon failure in Travis This file might have valuable content to identify a flaky test or any other problem.
* Wed Jul 05 2017 containers-bugownerAATTsuse.de- Commit d310c48 by Rafael Fernández López ereslibreAATTereslibre.es Remove `restart.txt` puma functionality Puma allows you to restart the server based on the `mtime` of a file `tmp/restart.txt`. We won\'t be using this functionality as the server runs in a container and we have other means for restarting it.
* Mon Jul 03 2017 containers-bugownerAATTsuse.de- Commit a187b77 by Vítor Avelino contactAATTvitoravelino.me ux/ui: added title to bootstrap button
* Fri Jun 30 2017 containers-bugownerAATTsuse.de- Commit bd67db1 by Flavio Castelli fcastelliAATTsuse.com Proxy support: code cleanup Apply the changes requested during the code review process. Signed-off-by: Flavio Castelli Commit 5bc7fd9 by Flavio Castelli fcastelliAATTsuse.com AutoYaST: handle proxy settings [bsc#1043538] Handle the proxy settings inside of the AutoYaST file when the user enabled them at the node level. This is fixes bsc#1043538 Signed-off-by: Flavio Castelli Commit d28c5ad by Flavio Castelli fcastelliAATTsuse.com Setup page: improve proxy settings This commit changes the proxy settings area of the setup page in the following ways:
* Introduce toggle switch that enables/disables proxy settings
* When the proxy settings are disabled hide the panel showing all their input fields.
* When the proxy settings are enabled show the panel showing all their input fields.
* Introduce toggle switch to enable/disable proxy settings at the node level. This is required to handle the `proxy:systemwide` pillar introduced by [this](https://github.com/kubic-project/salt/pull/107) PR. Signed-off-by: Flavio Castelli Commit e437f14 by Flavio Castelli fcastelliAATTsuse.com Handle removal of empty pillars Certain pillars (like the proxy related ones) can be blank. When that happens they must be removed from the `pillars` table of the database. This commit allows a user to effectively disable proxy related configurations by simply providing an empty value. The previous code would have ignored this change and left the old values instead. Signed-off-by: Flavio Castelli Commit a9b5441 by Flavio Castelli fcastelliAATTsuse.com Add support of proxy:systemwide pillar This pillar was introduced by [this](https://github.com/kubic-project/salt/pull/107) pull request but was never supported by Velum. Signed-off-by: Flavio Castelli
* Fri Jun 30 2017 containers-bugownerAATTsuse.de- Commit 652b0da by Vítor Avelino contactAATTvitoravelino.me ui: polishing and minor enhancements Most of changes were done based on wireframes: - General - Added missing Roboto Medium (used on panel titles for example) - Slightly larger line-height on `.panel-heading` to fit `.btn-sm` - Buttons on `.panel-heading` moved from `.btn-xs` to `.btn-sm` - Configuration - Added missing `.panel-title` (except for proxy settings to avoid conflict) - Overview page - Centered status icon to column name - Status text in the same column as icon - Aligned table data text in the middle - Wrong use of th instead of td on nodes table - Master checkbox aligned before role text - Changed update all nodes from button to link - Overview and Discovery - Added text when Pending Nodes panel is empty
* Fri Jun 30 2017 containers-bugownerAATTsuse.de- Commit b34d4ce by Maximilian Meister mmeisterAATTsuse.de hide minimum nodes alert (bsc#1046626) Signed-off-by: Maximilian Meister
* Fri Jun 30 2017 containers-bugownerAATTsuse.de- Commit 07795f3 by Rafael Fernández López ereslibreAATTereslibre.es Only restrict routes on production
* Thu Jun 29 2017 containers-bugownerAATTsuse.de- Commit ee0ee4e by Rafael Fernández López ereslibreAATTereslibre.es Add support for running Puma with SSL Puma will be executed using SSL. If the port provided is 80 (we will be serving autoyast with no safe transport in 1.0), then we only create the autoyast route, since we should not let the customers handle Velum through an insecure connection. Connect to the salt-api using TLS. Remove certificate details from the UI and related code. Fixes: bsc#1043570
* Thu Jun 29 2017 containers-bugownerAATTsuse.de- Commit 6054c65 by Miquel Sabaté Solà msabateAATTsuse.com Clarification on the discovery page Signed-off-by: Miquel Sabaté Solà
* Wed Jun 28 2017 containers-bugownerAATTsuse.de- Commit ae40fb3 by Vítor Avelino contactAATTvitoravelino.me Added \'cluster is too small\' warning The etcd configuration used in CaaSP 1.0 requires three total nodes to be present in the cluster (can be 1 master and two minions) - without this, etcd does not elect a leader, and bootstrap will fail. Warns the user if he tries to setup a cluster with only two nodes. See bsc#1040477
* Tue Jun 27 2017 containers-bugownerAATTsuse.de- Commit afb1d63 by Vítor Avelino contactAATTvitoravelino.me Fixed \'Reboot to update\' after click behavior After clicking \"Reboot to update\", fixed wrong button text and locked modal until admin node is rebooted. bsc#1045380, bsc#1045730
* Tue Jun 27 2017 containers-bugownerAATTsuse.de- Commit 4efd537 by Vítor Avelino contactAATTvitoravelino.me ui: discovery page layout enhancement
* Tue Jun 27 2017 containers-bugownerAATTsuse.de- Commit ebb8faa by Michal Jura mjuraAATTsuse.com Fix minions update status, bsc#1045734 We have to add and check update_status from minions equal to 3, which means that a highstate failed to apply to the node, which is a retryable upgrade failure, so we should allow users to re-trigger an upgrade, which we do by enabling the \"Update all Nodes\" button
* Fri Jun 23 2017 containers-bugownerAATTsuse.de- Commit a8ed511 by Graham Hayes graham.hayesAATTsuse.com bsc#1043592 Add support for accepting nodes Turning off `auto_accept` for salt means we need to have a mechanism to accept nodes on a node by node basis into the cluster. This adds a pending nodes table that will show all nodes that have not had their key accepted in salt yet.
* Fri Jun 23 2017 containers-bugownerAATTsuse.de- Commit 70049ca by Vítor Avelino contactAATTvitoravelino.me Missing improvements from #201 (add unassigned nodes)
* Thu Jun 22 2017 containers-bugownerAATTsuse.de- Commit 71840dd by Graham Hayes graham.hayesAATTsuse.com Add \'grains_refresh_every\' config minions
* Wed Jun 21 2017 containers-bugownerAATTsuse.de- Commit bcdecc9 by Vítor Avelino contactAATTvitoravelino.me Add unassigned nodes after cluster bootstrap After the initial selection of the cluster (discovery page), new nodes can eventually come up and user might want to add them. This PR makes that available with a dedicated page similar to the discovery one. https://trello.com/c/yzU3t5EC/229-5-ui-create-the-add-unassigned-nodes-to-the-cluster
* Wed Jun 21 2017 containers-bugownerAATTsuse.de- Commit 33506ea by Maximilian Meister mmeisterAATTsuse.de Gemfile.lock: downgrade a few gems the recent capybara/poltergeist update can live with lower versions of their deps as well partial revert of: 2a2125e8fc4ad66c7901ba6f0db7462c65056e41 Signed-off-by: Maximilian Meister
* Tue Jun 20 2017 containers-bugownerAATTsuse.de- Commit b1d0cd6 by Rafael Fernández López ereslibreAATTereslibre.es Remove phantomjs gem Add phantomjs to travis configuration
* Tue Jun 20 2017 containers-bugownerAATTsuse.de- Commit 1052024 by Vítor Avelino contactAATTvitoravelino.me Allows selection of nodes to bootstrap cluster Previously, all the nodes found were used to be part of the cluster. Now, user needs to select which ones they want to be used for it.
* Tue Jun 20 2017 containers-bugownerAATTsuse.de- Commit 2a2125e by Miquel Sabaté Solà msabateAATTsuse.com Fixed flaky tests - Some versions have been upgraded. - I\'ve taken the `spec/support/wait_for_events.rb` from Portus and used some of its methods here and there. - I\'ve also tweaked some tests. Thanks a lot to AATTvitoravelino and AATTereslibre for all the help for fixing this. Signed-off-by: Miquel Sabaté Solà
* Tue Jun 20 2017 containers-bugownerAATTsuse.de- Commit f86b8d6 by Rafael Fernández López ereslibreAATTereslibre.es When setting the highstate resolution, consider all highstates With the current logic, it could happen that during the same orchestration run the first highstate succeeds (setting the minion status to :applied as succeeded), while the second highstate fails (and the :applied status as succeeded will be kept). With this change we consider the latest highstate run is the most recent, and whatever it reports is what the final status of the minion will be.
* Fri Jun 16 2017 containers-bugownerAATTsuse.de- Commit aecdaa5 by Vítor Avelino contactAATTvitoravelino.me Fixed regression introduced when changed dashboard layout For 1.0 we changed the layout of overview page. During that process, I accidentally changed this piece of code without reminding that the discovery page also depended on this. Now it separates different behavior depending on the page.
* Wed Jun 14 2017 containers-bugownerAATTsuse.de- Commit 886e7ae by Graham Hayes graham.hayesAATTsuse.com Add \"update\" button to velum.
* Adds a button that will trigger the update orchestration run, when there is nodes that can be updated.
* Updates the status column to show status of updates.
* Updated the Salt Event listener to update status of minions marked as \"failed\" if another highstate succeeds
* Wed Jun 14 2017 containers-bugownerAATTsuse.de- Commit e1aaf8b by Miquel Sabaté Solà msabateAATTsuse.com Don\'t be so strict on some pillar fields Signed-off-by: Miquel Sabaté Solà
* Wed Jun 14 2017 containers-bugownerAATTsuse.de- Commit 9d23ad4 by Valentin Rothberg vrothbergAATTsuse.com registration: allow only one admin user The administration dashboard allows to create as many admin users as desired, which is a potential security threat as Velum does not have any user management. Besides, an admin cannot see how many other users exist. Avoid creating more than one admin user by not showing the `Create an account` button when a user already exists. `users/sign_up` will further redirect to the `root_path` and print an alert. Notice that this is only an intermediate solution until multiuser support and RBAC is implemented. Until then, a password reset must be done by connecting to the machine (manually or via a script). See bsc#1040831 Signed-off-by: Valentin Rothberg
* Wed Jun 14 2017 containers-bugownerAATTsuse.de- Commit 1212762 by Rafael Fernández López ereslibreAATTereslibre.es Check that result isn\'t `nil` before checking result contents We are checking that the referenced file exists by performing a rudimentary check. However, the result might be completely empty under certain circumstances, making our check fail (`include?` called on `nil` will fail). This fixes the 500 error, but we still need to find out why the salt side of things is failing at all to retrieve the contents of the file. Fixes: bsc#1043843
* Tue Jun 13 2017 containers-bugownerAATTsuse.de- Commit 476b5d5 by Vítor Avelino contactAATTvitoravelino.me dashboard: updated layout based on wireframes
* Mon Jun 12 2017 containers-bugownerAATTsuse.de- Commit de90154 by Miquel Sabaté Solà msabateAATTsuse.com Added support for detection & reboot on the admin node Signed-off-by: Miquel Sabaté Solà
* Fri Jun 09 2017 containers-bugownerAATTsuse.de- Commit 44bc253 by Rafael Fernández López ereslibreAATTereslibre.es Remove the development environment and move it to its own repo The development environment can be used to hack both on Salt and Velum. Clean up references to the development environment and move all that to its dedicated repository.
* Thu Jun 08 2017 containers-bugownerAATTsuse.de- Commit 039ff47 by Flavio Castelli fcastelliAATTsuse.com Improve initial cluster config page [bsc#1040367] This is not the definitive design of the initial cluster configuration page, it\'s just an attempt to quickly improve it. The different forms have been grouped together based on the area they are related to. Also hover help tips have been added to the forms that might be harder to understand. This helps to mitigate issue bsc#1040367 Signed-off-by: Flavio Castelli
* Wed Jun 07 2017 containers-bugownerAATTsuse.de- Commit 92f5a04 by Vítor Avelino contactAATTvitoravelino.me signup: added proper email validation I had to add checkValidity() to force a manual check (the same as when form is submitted) to avoid the weird default suggestions and enforce the custom message. See bsc#1041940
* Tue Jun 06 2017 containers-bugownerAATTsuse.de- Commit 84e0ea1 by Jordi Massaguer Pla jmassaguerplaAATTsuse.de fix version in package to be a sequence Let\'s use the git revision number in the version so that zypper will see each new revision as an update. Otherwise, if we only use the commit number, zypper gets confused- Commit 1f38f52 by Rafael Fernández López ereslibreAATTereslibre.es Fix `minion` -> `worker` terminology Internally we refer to `minion` as a salt minion generically, but when it comes to assign roles (that will be shown to the user), fix terminology and use `worker` instead of `minion`. This fixes bsc#1032917- Commit 2594fe9 by Miquel Sabaté Solà msabateAATTsuse.com Added Docker HTTP proxy settings See kubic-project/salt#66 See bsc#1036627 Signed-off-by: Miquel Sabaté Solà - Commit e9673d5 by Miquel Sabaté Solà msabateAATTsuse.com Don\'t allow users to download the kubeconfig when bootstrapping Unless all the nodes are up and ready, don\'t allow users to download the kubeconfig file. See bsc#1041806 Signed-off-by: Miquel Sabaté Solà
* Fri Jun 02 2017 containers-bugownerAATTsuse.de- Commit cb0d67b by Kiall Mac Innes kiallAATTmacinnes.ie Fix compatibility with salt 2016.11.4
* Fri Jun 02 2017 containers-bugownerAATTsuse.de- Commit 4c0e315 by Flavio Castelli fcastelliAATTsuse.com UI: show salt ID instead of rails\' internal ID [bsc#1041789] Change the UI to show the salt ID (which now is set to be equal to the machine-id) instead of showing Rails\' internal ID. This is required to fix bsc#1041789 Signed-off-by: Flavio Castelli
* Wed May 31 2017 containers-bugownerAATTsuse.de- Commit 028069e by Miquel Sabaté Solà msabateAATTsuse.com signup: added proper validation message for the password fields See bsc#1041905 Signed-off-by: Miquel Sabaté Solà
* Tue May 30 2017 containers-bugownerAATTsuse.de- Commit 9d7b06e by Vítor Avelino contactAATTvitoravelino.me Fixed footer overlapping elements If a page didn\'t fit users screen height, the footer overlaps any element since its position is absolute to .container element. An example of that was the \"Next\" button on initial configuration form. See bsc#1041440
* Mon May 29 2017 containers-bugownerAATTsuse.de- Commit a97129d by Miquel Sabaté Solà msabateAATTsuse.com- setup: Block non-alphabetical characters-- I\'ve also fixed a redirect.-- See bsc#1031475-- Signed-off-by: Miquel Sabaté Solà
* Thu May 25 2017 containers-bugownerAATTsuse.de- Commit f5f660b by Rafael Fernández López ereslibreAATTereslibre.es- Remove Dockerfile that was for development purposes-- We already have the caasp-devenv and don\'t really need this Dockerfile,- however it would be good to reintroduce it based on opensuse in the future.
* Thu May 25 2017 containers-bugownerAATTsuse.de- Commit ac24623 by Miquel Sabaté Solà msabateAATTsuse.com- autoyast: added docker-- See bsc#1040579-- Signed-off-by: Miquel Sabaté Solà
* Fri May 19 2017 containers-bugownerAATTsuse.de- Commit 1a388b6 by Rafael Fernández López ereslibreAATTereslibre.es- Ask for system activations and retrieve the regcode from this service-- We will identify an active registration code by the identifier and not the- name of the product.-- Fixes: bsc#1038074
* Thu May 18 2017 containers-bugownerAATTsuse.de- Commit 98effdd by Rafael Fernández López ereslibreAATTereslibre.es- Fix unit tests
* Thu May 18 2017 containers-bugownerAATTsuse.de- Commit 22ab346 by Rafael Fernández López ereslibreAATTereslibre.es- Add setting for the main apiserver hostname-- Fixes bsc#1039437
* Wed May 17 2017 containers-bugownerAATTsuse.de- Fix patterns on autoyast (Fixes: bsc#1039432)
* Tue May 16 2017 containers-bugownerAATTsuse.de- Use colons as nesting instead of dots
* Tue May 16 2017 containers-bugownerAATTsuse.de- Add support for libvirt in Vagrantfile
* Fri May 12 2017 containers-bugownerAATTsuse.de- Disable rubocop inline rescue warning
* Fri May 12 2017 containers-bugownerAATTsuse.de- Fix documentation typos
* Fri May 12 2017 containers-bugownerAATTsuse.de- autoyast: create /var/lib/docker partition (bsc#1034942)
* Fri May 12 2017 containers-bugownerAATTsuse.de- Propagate ssh public key from dashboard to workers, bsc#1030876
* Fri May 12 2017 containers-bugownerAATTsuse.de- Remove GPLv3+ license from specific files defaulting to project\'s license
* Fri May 12 2017 containers-bugownerAATTsuse.de- Add debugging- Add salt-api container configuration- Use the production schema to create the database at the beginning- Make the development environment as close as possible to production
* Tue May 09 2017 containers-bugownerAATTsuse.de- fix bsc#1036724
* Mon May 08 2017 containers-bugownerAATTsuse.de- Revert \"Do not bundle install in velum\'s source code location\"- Revert \"Fix path typo\"- Revert \"Set .bundle/config location by stripping the buildroot automatically\"
* Mon May 08 2017 containers-bugownerAATTsuse.de- packaging: use /usr/share instead of /usr/local
* Mon May 08 2017 containers-bugownerAATTsuse.de- Set .bundle/config location by stripping the buildroot automatically added when installing with --path
* Mon May 08 2017 containers-bugownerAATTsuse.de- Gemfile: fix ENV
* Thu May 04 2017 containers-bugownerAATTsuse.de- Fix path typo
* Thu May 04 2017 containers-bugownerAATTsuse.de- Do not bundle install in velum\'s source code location
* Wed May 03 2017 containers-bugownerAATTsuse.de- discovery: don\'t cache AJAX requests (fix bsc#1031742)
* Thu Apr 27 2017 containers-bugownerAATTsuse.de- docs: added internal research on salt job callbacks
* Thu Apr 27 2017 containers-bugownerAATTsuse.de- Fixed the node count on the discovery page (bsc#1036040)
* Tue Apr 25 2017 containers-bugownerAATTsuse.de- Prevent exception in console during initial bootstrap
* Tue Apr 25 2017 containers-bugownerAATTsuse.de- Update velum-secrets host path on development
* Tue Apr 25 2017 containers-bugownerAATTsuse.de- Add seeding depending on environment
* Tue Apr 25 2017 containers-bugownerAATTsuse.de- Remove dashboard pillar seed
* Mon Apr 24 2017 containers-bugownerAATTsuse.de- Explicitly ask for the dashboard location (bsc#1032461 and bsc#1031605)
* Mon Apr 24 2017 containers-bugownerAATTsuse.de- we always need to specify a VELUM_SECRETS_DIR
* Mon Apr 24 2017 containers-bugownerAATTsuse.de- enhance secrets method and use /var/lib/misc to store the secrets- Make rubocop happy- Store and retrieve secret_key_base from disk file
* Fri Apr 21 2017 containers-bugownerAATTsuse.de- Cleaning folders that can be problematic
* Wed Apr 19 2017 containers-bugownerAATTsuse.de- Added NTP configuration for nodes in autoyast.xml
* Wed Apr 19 2017 containers-bugownerAATTsuse.de- clanup: also don\'t persist the server.pid file- init: define VELUM_PORT earlier
* Wed Apr 19 2017 containers-bugownerAATTsuse.de- manifest-templates: always use the latest images
* Wed Apr 19 2017 containers-bugownerAATTsuse.de- packaging: rebase patch
* Wed Apr 19 2017 containers-bugownerAATTsuse.de- Do not assume minion_id is hostname/fqdn
* Tue Apr 18 2017 containers-bugownerAATTsuse.de- Forcing a user to specify a SALT_DIR
* Thu Apr 13 2017 containers-bugownerAATTsuse.de- Use the same init script for development env as we do in production
* Thu Apr 13 2017 containers-bugownerAATTsuse.de- Improve bootstrap documentation (bsc#1031685)
* Thu Apr 13 2017 containers-bugownerAATTsuse.de- fix bsc#1031682
* Thu Apr 13 2017 containers-bugownerAATTsuse.de- Update manifests.- Bug 1031352 - Do not expose mariadb to the world
* Tue Apr 11 2017 containers-bugownerAATTsuse.de- Fix kubelet arguments for version >= 1.6.1
* Fri Mar 31 2017 containers-bugownerAATTsuse.de- Request the ca certificate directly to the CA machine (container) (bsc#1031478)
* Tue Mar 28 2017 containers-bugownerAATTsuse.de- bootstrap: don\'t allow empty bootstrap in the back-end
* Mon Mar 27 2017 containers-bugownerAATTsuse.de- setup: enable select field for the country codes
* Mon Mar 27 2017 containers-bugownerAATTsuse.de- discovery: don\'t allow empty bootstrap
* Mon Mar 27 2017 containers-bugownerAATTsuse.de- Set \"dashboard\" Pillar to the request.host value
* Mon Mar 27 2017 containers-bugownerAATTsuse.de- Fix spacings
* Mon Mar 27 2017 containers-bugownerAATTsuse.de- Add netsetup=dhcp documentation
* Mon Mar 27 2017 containers-bugownerAATTsuse.de- packaging: rename patch
* Fri Mar 24 2017 containers-bugownerAATTsuse.de- Set salt master configuration on the minion on a script
* Fri Mar 24 2017 containers-bugownerAATTsuse.de- Remove SECRET_KEY_BASE
* Fri Mar 24 2017 containers-bugownerAATTsuse.de- setup: validate fields in the front end
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- setup: don\'t allow empty initial configuration
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Fix rubocop warning
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- add init for containers
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Remove `SKIP_MIGRATION` envvar
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Remove port from puma configuration
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Fixed the production environment
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Typo fixes
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Retrieve host from request
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Update coverage and codeclimate table in README.md
* Thu Mar 23 2017 containers-bugownerAATTsuse.de- Add AutoYaST profile serving
* Fri Mar 17 2017 containers-bugownerAATTsuse.de- Provide kubectl configuration
* Fri Mar 17 2017 containers-bugownerAATTsuse.de- provide hints about the initial configuration
* Wed Mar 15 2017 containers-bugownerAATTsuse.de- Change \"minions\" to \"workers\"
* Wed Mar 15 2017 containers-bugownerAATTsuse.de- Add bootstrap javascript to make close buttons on alerts work- ui: styling enhancements
* Wed Mar 15 2017 containers-bugownerAATTsuse.de- Show message about unassigned minions in monitoring page
* Mon Mar 13 2017 mmeisterAATTsuse.com- Implementing a setup controller to guide the user throught the setup- Overhaul of the UI- Drop turbolink dependencies. They have been removed with commit: 6758b5ac35db111393339a5f4916ae28f4d794c3
* Thu Mar 09 2017 jmassaguerplaAATTsuse.com- Disable the service so we don\'t get accidentally rebuilds on our final repos- add 0_change_versions.gem.patch: This relaxes the requirement on recursive-open-struct to ~>1.0.0 instead of =1.0.0 so that we can use version 1.0.2 which is the latest version upstream and so the one that we are packaging. This change has been made also upstream in the master branch of kubeclient, so that this is a backport which should be removed when we update kubeclient
* Fri Feb 10 2017 mmeisterAATTsuse.com- Initial packaging of velum