* feat(api): gitlab registry type
* feat(registries): early support for gitlab registries
* feat(app): registry service selector
* feat(registry): gitlab support : list repositories and tags - remove features missing
* feat(registry): gitlab registry remove features
* feat(registry): gitlab switch to registry V2 API for repositories and tags
* feat(api): use development extension binary
* fix(registry): avoid 401 on gitlab retrieve to disconnect the user
* feat(registry): gitlab browse projects without extension
* style(app): code cleaning
* refactor(app): PR review changes + refactor on types
* fix(gitlab): remove gitlab info from registrymanagementconfig and force gitlab type
* style(api): go fmt
* feat(api): update APIVersion and ExtensionDefinitionsURL
* fix(api): fix invalid RM extension URL
* feat(registry): PAT scope help
* feat(registry): defaults on registry creation
* style(registry-creation): update layout and text for Gitlab registry
* feat(registry-creation): update gitlab notice
Currently we are using RegExp `/\=(.+)/` to catch key-value
of environment variables, which could not match empty-value
environment variables such as `KEY=`.
This commit will change the RegExp to `/\=(.*)/`, which
matches the empty values.
* feat(containers): added support for port range mappings when deploying containers
* feat(containers): added placeholders to port publishing input fields
* feat(containers): added a tooltip to the manual network port publishing
* feat(containers): improved the code consistency
* fix(registry): fetch datatable details on page/filter/order state change instead of fetching all data on first load
* fix(registry): fetch tags datatable details on state change instead of fetching all data on first load
* fix(registry): add pagination support for tags + loading display on data load
* fix(registry): debounce on text filter to avoid querying transient matching values
* refactor(registry): rebase on latest develop
* feat(registries): background tags and optimisation -- need code cleanup and for-await-of to cancel on page leave
* refactor(registry-management): code cleanup
* feat(registry): most optimized version -- need fix for add/retag
* fix(registry): addTag working without page reload
* fix(registry): retag working without reload
* fix(registry): remove tag working without reload
* fix(registry): remove repository working with latest changes
* fix(registry): disable cache on firefox
* feat(registry): use jquery for all 'most used' manifests requests
* feat(registry): retag with progression + rewrite manifest REST service to jquery
* fix(registry): remove forgotten DI
* fix(registry): pagination on repository details
* refactor(registry): info message + hidding images count until fetch has been done
* fix(registry): fix selection reset deleting selectAll function and not resetting status
* fix(registry): resetSelection was trying to set value on a getter
* fix(registry): tags were dropped when too much tags were impacted by a tag removal
* fix(registry): firefox add tag + progression
* refactor(registry): rewording of elements
* style(registry): add space between buttons and texts in status elements
* fix(registry): cancelling a retag/delete action was not removing the status panel
* fix(registry): tags count of empty repositories
* feat(registry): reload page on action cancel to avoid desync
* feat(registry): uncancellable modal on long operations
* feat(registry): modal now closes on error + modal message improvement
* feat(registries): remove empty repositories from the list
* fix(registry): various bugfixes
* feat(registry): independant timer on async actions + modal fix
* feat(container-details): add entrypoint to container details view
* feat(container-details): restore file from develop branch to bring back original indentation
* feat(container-details): add entrypoint to container details view
* feat(services): rollback service capability
* refactor(services): notification reword
Co-Authored-By: William <william.conquest@portainer.io>
* refactor(services): remove TODO comment + add note on rollback capability
* fix(services): service update rpc error version out of sync
* feat(services): confirmation modal on rollback
* feat(services): rpc error no previous spec message
* feat(swarmvisualizer): add labels display under node info
* feat(swarmvisualizer): fix css
* add toggle to display node labels
* feat(swarmvisualizer): rename filters section + fix display when label has no value
* feat(swarmvisualizer): retrieve state from local storage for node labels display toggle
* fix(datatables): allow selecting range using shift (#344)
* feat(datatables): more intuitive batch select behaviour
* feat(datatables): add overridable function called on selection change
* refactor(datatables): remove custom selectAll on Generic-extending Controllers
* fix(datatables): stored state data retrieval on Generic-extanding datatables controllers
* refactor(datatables): remove code duplication between GenericController and extending controllers
* fix(app): wrapper for UI refresh trigger with async/await
* fix(async): $async wrapper now accepts functions with params
* fix(async): $async should return a promise to be chained with ES5 .then() style
* fix(async): $async with multiple params was not working
* refactor(app): wrap all async functions with $async
* docs(async): add link to async wrapper documentation
* fix(networks): disable removing predefined networks (#1838)
* fix(networks): disable select all for predefined networks (#1838)
* fix(networks): do not allow delete in network-details & use constant (#1838)
* refactor(stacks): set newstack state as a child state of stacks
* fix(docker): add check on docker states for endpoint
* refactor(app): remove redirect notification
* fix(container): rename old container only if exist
* fix(container): remove new container only if created
* style(container): fix typo
Co-Authored-By: chiptus <chiptus@users.noreply.github.com>
* fix(network-creation): force overlay network creation on manager node
* fix(app): fix function override
* fix(app): use portainerAgentManagerOperation in interceptor
* feat(settings): add the capability to enable/disable the host management features
* feat(settings): remove the validation of EnableHostManagementFeatures in frontend
* feat(api): disable schedules API when HostManagementFeatures is false + DB migration
* style(settings): update host management settings tooltip
* refacot(schedules): update DBVersion to 15
* chore(log-viewer): add the ability to use`since` parameter #1942https://github.com/portainer/portainer/issues/1942#issuecomment-430246378
* chore(log-viewer): change lineCount to 100 #1942https://github.com/portainer/portainer/issues/1942#issuecomment-430246378
* fix(log-viewer): js syntax typo for `;` and `'`
forget to lint the code, reported by codeclimate
* fix(log-viewer): use mementjs to format timestamp
1. use moment lib instead of define a function in filter.js(not the right place for this function, removed)
2. set sinceTimestamp init value to `24 hours ago`, as we just need to focus on the relative latest logs after the log-viewer loading, not all the logs(to speedup the process)
3. use moment().unix() to convert the `sinceTimestamp` to local unix timestamp(not utc)
* chore(log-viewer): add the ability to select the datetime for `since`
* chore(log-viewer): add the ability to fetch logs from specific time
* feat(back): saved data in snapshot
* feat(endpoints): adding interceptors to retrieve saved data on offline endpoints
* feat(endpoints): offline dashboard working - need tests on offline views
* refactor(endpoints): interceptors cleaning and saving/loading offline endpoints data in/from localstorage
* feat(endpoints): browsing offline endpoints
* feat(endpoints): removing all the link in offline mode - sidebar not working when switching between off and on modes w/ stateManager logic
* feat(endpoints): endpoint status detection in real time
* fix(endpoints): offline swarm endpoint are not accessible anymore
* fix(endpoints): refactor message + disable offline browsing for an endpoint when no snapshot is available for it
* fix(endpoints): adding timeout and enabling loading bar for offline requests
* fix(endpoints): trying to access a down endpoint wont remove sidebar items if it fails
* feat(endpoints): disable checkboxes on offline views for offline mode
* feat(endpoints): updating endpoint status when detecting a change
* refactor(host): moved offline status panel from engine view to new host view
* fix(endpoints): missing endpoint update on ping from home view
* fix(api): rework EndpointUpdate operation
* refactor(offline): moved endpoint status to EndpointProvider and refactor the status-changed detection
* fix(offline): moved status detection to callback on views -> prevent displaying the offline message when endpoint is back online on view change
* fix(offline): offline message is now displayed online when browsing an offline endpoint
* fix(offline): sidebar updates correctly on endpoint status change
* fix(offline): offline panel not displayed and hidden on online mode
* refactor(offline): rework of OfflineMode management
* refactor(offline): extract information-panel for offlineMode into a component
* refactor(offline): remove redundant binding of informationPanel + endpointStatusInterceptor patter as service
* refactor(interceptors): moved interceptors pattern to service pattern
* feat(stacks): prevent inspection of a stack in offline mode
* feat(host): hide devices/disk panels in offline mode
* feat(host): disable browse action in offline mode
* refactor(home): remove comments
* feat(jobs): adding the ability to run scripts on endpoints
fix(job): click on containerId in JobsDatatable redirects to container's logs
refactor(job): remove the jobs datatable settings + texts changes on JobCreation view
fix(jobs): jobs payloads are now following API rules and case
feat(jobs): adding the capability to run scripts on hosts
* feat(jobs): adding the ability to purge jobs containers
* refactor(job): apply review changes
* feat(job-creation): store image name in local storage
* feat(host): disable job exec link in non-agent Swarm setup
* feat(host): only display execute job in agent setups or standalone
* feat(job): job execution overhaul
* docs(swagger): update EndpointJob documentation
* feat(agent): get agent's version from ping
* feat(agent): add version to api url
* feat(agent): query agent with api version
* feat(agent): rename agent api version name on state
* feat(agent): disable feature based on agent's api version
* style(agent): rename ping rest service + remove whitespaces
* style(state): remove whitespace
* style(agent): add whitespace
* fix(agent): remove check for error status 403
* refactor(agent): rename ping file name
* refactor(agent): move old services to v1 folder
* refactor(agent): turn ping service to usual pattern
* refactor(agent): change version to a global variable
* refactor(agent): move ping to version2
* refactor(agent): restore ping to use root ping
* fix(volumes): add volumeID to browse api path
* feat(volume): add upload button to volume browser
* fix(containers): creating a container with default runtime let the docker daemon assume the correct value
* refactor(containers): implementation simplification of default runtime value
* containersDatable: add containers name if error on executeActionOnContainerList
* Update containersDatatableActionsController.js
* Update containersDatatableActionsController.js
* Initial pass at adding webhook controller and routes
* Moving some objects around
* Cleaning up comments
* Fixing syntax, switching to using the docker sdk over building an http client
* Adding delete and list functionality
* Updating the handler to use the correct permissions. Updating some comments
* Fixing some comments
* Code cleanup per pull request comments
* Cleanup per PR feedback. Syntax error fix
* Initial creation of webhook app code
* Moving ClientFactory creation out of handler code and instead using the one created by the main process. Removing webhookInspect method and updating the list function to use json filters
* Delete now works on the webhook ID vs service ID
* WIP - Service creates a webhook. Display will show an existing webhook URL.
* Adding the webhook field to the service view. There is now the ability to add or remove a webhook from a service
* Moving all api calls to be webhooks vs webhook
* Code cleanup. Moving all api calls to be webhooks vs webhook
* More conversion of webhook to webhooks?
* Moving UI elements around. Starting function for copying to clipboard
* Finalizing function for copying to clipboard. Adding button that calls function and copies webhook to clipboard.
* Fixing UI issues. Hiding field entirely when there is no webhook
* Moving URL crafting to a helper method. The edit pane for service now creates/deletes webhooks immidiately.
* style(service-details): update webhook line
* feat(api): strip sha when updating an image via the update webhook
* Fixing up some copy. Only displying the port if it is not http or https
* Fixing tooltip copy. Setting the forceupdate to be true to require an update to occur
* Fixing code climate errors
* Adding WebhookType field and setting to ServiceWebhook for new webhooks. Renaming ServiceID to resourceID so future work can add new types of webhooks in other resource areas.
* Adding the webhook type to the payload to support more types of webhooks in the future. Setting the type correctly when creating one for a service
* feat(webhooks): changes related to webhook management
* API code cleanup, removing unneeded functions, and updating validation logic
* Incorrectly ignoring the error that the webhook did not exist
* Re-adding missing error handling. Changing error response to be a 404 vs 500 when token can't find an object
* fix(webhooks): close Docker client after service webhook execution
* feat(api): add motd handler
* feat(app): add the motd api layer
* feat(motd): display motd and add the ability to dismiss information messages
* style(home): relocate important message before info01
* feat(api): silently fail when an error occurs during motd retrieval
* fix(services): replicas numbers display is now correct with constraints and down nodes
* refactor(helpers): constraint helper has less complexity
* feat(services): constraints on node/engine labels are now supported
* refactor(helpers): ConstraintsHelper - remove regex patterns and improve code lisibility
* refactor(helpers): rework matchesConstraint() for better code lisibility and lodash find() instead for IE compatibility
* #960 feat(UAC): change ownership to admins for externally created ressources
* feat(UAC): change ownership to admins for externally created resources
Deprecated AdministratorsOnly js and go backend
* #960 feat(UAC): remove AdministratorsOnly property and minor GUI fixes
Update swagger definition changing AdministratorsOnly to Public
* #960 feat(UAC): fix create resource with access control data
* #960 feat(UAC): authorization of non-admin users for restricted operations
On stacks, containers networks, services , tasks and volumes.
* #960 feat(UAC): database migration to version 14
The administrator resources are deleted and Public resources are now managed by admins
* #960 feat(UAC): small fixes from PR #2137
* #960 feat(UAC): improve the readability of the source code
* feat(UAC) fix displayed ownership for Swarm related resources (#960)
* refactor(container-creation): change order of container creation steps
* refactor(container-creation): remove nested methods
* fix(container-creation): skip actions if old container missing
* fix(container-creation): reject if user is not authorized
* fix(container-creation): remove rejection on invalid form
* refactor(container-creation): start container after duplicate
* fix(container-creation): add form validation error message
* fix(container-creation): pass correct id to create resource control
* fix(container-creation): set action in progress after confirmation
* feat(network-creation): macvlan driver for swarm
* refactor(network-creation): layout rework to make it simpler with MACVLAN and keep it consistent with other drivers
* fix(network-creation): MACVLAN - parent network card is now properly saved, names are not prefixed anymore and the --attachable option is now supported
* refactor(network-creation): PR macvlan review - rework of macvlan view + code optimisation
* fix(network-creation): disable attachable and internal options on macvlan config creation
* fix(home): show cpu/mem for swarm
* fix(home): add nodes data to snapshot
* fix(dashboard): get cpus/mem from snapshot
* refactor(home): remove temp variable
* feat(container-creation): container add/drop capabilities on creation
* feat(container-creation): capabilities are now loaded on edit/duplicate/update
* feat(volume-creation): NFS support for volume creation - layout
* feat(volume-creation): NFS support for volume creation
* fix(volume-creation): NFS style, display and check on submit
* refactor(volume-creation): remove useless controller + refactor var naming
* refactor(volume-creation): NFS wording, help and style
* fix(style): wrap long text in tables cells (#1920)
* fix(style): <code> tags are now wrapped correctly (PR #2052)
* fix(style): revert #1770 style-related content and apply nowrap on datatables (PR#2052)
* feat(services): add the ability to pull latest image when updating a service
* feat(services): update version header value
* refactor(services): remove TODO
* feat(services): rollback version header value to 1.29
* feat(stacks): add the ability to migrate stacks to another endpoint
* feat(stack-details): do not redirect to alternate endpoint after migration
* fix(api): fix merge conflicts
* feat(stack-details): add a modal to confirm stack migration
* fix(api): add an authenticated access policy to the websocket endpoint
* refactor(api): centralize EndpointAccess validation
* feat(api): validate id query parameter for the /websocket/exec endpoint
REST call to stop/restart a container overrides the default stop timeout (before kill) with hardcoded 5 seconds.
Containers already have a default stop timeout handled by the engine API (https://github.com/moby/moby/blob/master/client/container_stop.go).
With this hardcoded 5 seconds, the containers get killed after 5 seconds even if they define a custom greater stop timeout.
Another solution would be to not hardcode the 5 seconds but rather use a global editable setting.
* fix(container): Use first network's Mac address by default
* fix(container): Connect additional networks to container after creation
* fix(container): Remove warning message
* feat(image-details): display image layer depth and sort by it by default (#1706)
* refactor(image-details): rename 'Depth' to 'Order' in image layers table
* refactor(image-details): sort image layers from the bottom to the top one
* feat(log-viewer): use only one switch to manage collection/autoscroll
* feat(log-viewer): add the ability to clear selection
* style(log-viewer): update unselect button design
William