portainer/api/http/handler/stacks/handler.go

247 lines
8.2 KiB
Go
Raw Normal View History

package stacks
import (
"context"
"fmt"
"net/http"
"strings"
"sync"
"github.com/docker/docker/api/types"
"github.com/gorilla/mux"
"github.com/pkg/errors"
httperror "github.com/portainer/libhttp/error"
feat(compose): add docker-compose wrapper (#4713) * feat(compose): add docker-compose wrapper ce-187 * fix(compose): pick compose implementation upon startup * Add static compose build for linux * Fix wget * Fix platofrm specific docker-compose download * Keep amd64 architecture as download parameter * Add tmp folder for docker-compose * fix: line endings * add proxy server * logs * Proxy * Add lite transport for compose * Fix local deployment * refactor: pass proxyManager by ref * fix: string conversion * refactor: compose wrapper remove unused code * fix: tests * Add edge * Fix merge issue * refactor: remove unused code * Move server to proxy implementation * Cleanup wrapper and manager * feat: pass max supported compose syntax version with each endpoint * fix: pick compose syntax version * fix: store wrapper version in portainer * Get and show composeSyntaxMaxVersion at stack creation screen * Get and show composeSyntaxMaxVersion at stack editor screen * refactor: proxy server * Fix used tmp * Bump docker-compose to 1.28.0 * remove message for docker compose limitation * fix: markup typo * Rollback docker compose to 1.27.4 * * attempt to fix the windows build issue * * attempt to debug grunt issue * * use console log in grunt file * fix: try to fix windows build by removing indirect deps from go.mod * Remove tmp folder * Remove builder stage * feat(build/windows): add git for Docker Compose * feat(build/windows): add git for Docker Compose * feat(build/windows): add git for Docker Compose * feat(build/windows): add git for Docker Compose * feat(build/windows): add git for Docker Compose * feat(build/windows): add git for Docker Compose - fixed verbose output * refactor: renames * fix(stack): get endpoint by EndpointProvider * fix(stack): use margin to add space between line instead of using br tag Co-authored-by: Stéphane Busso <stephane.busso@gmail.com> Co-authored-by: Simon Meng <simon.meng@portainer.io> Co-authored-by: yi-portainer <yi.chen@portainer.io> Co-authored-by: Steven Kang <skan070@gmail.com>
2021-01-25 19:16:53 +00:00
portainer "github.com/portainer/portainer/api"
chore(store) EE-1981: Refactor/store/error checking, and other refactoring (#6173) * use the Store interface IsErrObjectNotFound() to avoid revealing internal errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * what happens when you extract the datastore interfaces into their own package Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Start renaming Storage methods Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the boltdb specific code from the Portainer storage code (example, the others need the same) Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extract bolt.Tx from datastore code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * minimise imports by putting moving the struct definition into the file that needs the Service imports Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extraction of boltdb.Tx Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the use of bucket.SetSequence Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * almost done - just endpoint.Synchonise :/ Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * so, endpoint.Synchonize looks hard, but i can't find where we use it, so 'delete first refactoring' Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix test compile errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * test compile fixes after rebase Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix a mis-remembering I had wrt deserialisation - last time i used AnyData - jsoniter's bindTo looks interesting for the same reason Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * set us up to make the connection an interface Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make the db connection a datastore interface, and separate out our datastore services from the bolt ones Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * rename methods to something less oltdb internals specific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * these errors are not boltdb secific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * start using the db-backend factory method too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * export boltdb raw in case we can't export from the service layer Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add a raw export from boltdb to yaml for broken db's, and an export services to yaml in backup Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add the version info by hand for now Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * actually, the export from services can be fully typed - its the import that needs to do more work Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * redo raw export, and make import capable of using it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add DockerHub Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * migration from anything older than v1.21.0 has been broken for quite a while, deleting the un-tested code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix go test ./... again Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * my goland wasn't setup to gofmt Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the two extremely dubious migration tests down into store, so they can use the test store code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * the migrator is now free of boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * reverse goland overzealous replcement of internal with boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more undo over-zealous goland internal->boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * yay, now bolt is only mentioned inside the api/database/ dir Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * and this might be the last of the boltdb references? Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add todo Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the store code into a separate module too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * don't need the fileService in boltdb anymore Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use IsErrObjectNotFound() Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use a string to select what database backend we use Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make isNew store an ephemeral bool that doesn't stay true after we've initialised it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the import.json wip to a separate file so its more obvious - we'll be using it for testing, emergency fixups, and in the next part of the store work, when we improve migrations and data model lifecycles Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * undo vscode formatting html Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix app templates symbol (#6221) * feat(webhook) EE-2125 send registry auth haeder when update swarms service via webhook (#6220) * feat(webhook) EE-2125 add some helpers to registry utils * feat(webhook) EE-2125 persist registryID when creating a webhook * feat(webhook) EE-2125 send registry auth header when executing a webhook * feat(webhook) EE-2125 send registryID to backend when creating a service with webhook * feat(webhook) EE-2125 use the initial registry ID to create webhook on editing service screen * feat(webhook) EE-2125 update webhook when update registry * feat(webhook) EE-2125 add endpoint of update webhook * feat(webhook) EE-2125 code cleanup * feat(webhook) EE-2125 fix a typo * feat(webhook) EE-2125 fix circle import issue with unit test Co-authored-by: Simon Meng <simon.meng@portainer.io> * fix(kubeconfig): show kubeconfig download button for non admin users [EE-2123] (#6204) Co-authored-by: Simon Meng <simon.meng@portainer.io> * fix data-cy for k8s cluster menu (#6226) LGTM * feat(stack): make stack created from app template editable EE-1941 (#6104) feat(stack): make stack from app template editable * fix(container):disable Duplicate/Edit button when the container is portainer (#6223) * fix/ee-1909/show-pull-image-error (#6195) Co-authored-by: sunportainer <ericsun@SG1.local> * feat(cy): add data-cy to helm install button (#6241) * feat(cy): add data-cy to add registry button (#6242) * refactor(app): convert root folder files to es6 (#4159) * refactor(app): duplicate constants as es6 exports (#4158) * fix(docker): provide workaround to save network name variable (#6080) * fix/EE-1862/unable-to-stop-or-remove-stack workaround for var without default value in yaml file * fix/EE-1862/unable-to-stop-or-remove-stack check yaml file * fixed func and var names * wrapper error and used bool for stringset * UT case for createNetworkEnvFile * UT case for %s=%s * powerful StringSet * wrapper error for extract network name * wrapper all the return err * store more env * put to env file * make default value None * feat: gzip static resources (#6258) * fix(ssl)//handle --sslcert and --sslkey ee-2106 (#6203) * fix/ee-2106/handle-sslcert-sslkey Co-authored-by: sunportainer <ericsun@SG1.local> * fix(server):support disable https only ee-2068 (#6232) * fix/ee-2068/disable-forcely-https * feat(store): implement store tests EE-2112 (#6224) * add store tests * add some more tests * Update missing helm user repo methods * remove redundant comments * add webhook export * update webhooks * use the Store interface IsErrObjectNotFound() to avoid revealing internal errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * what happens when you extract the datastore interfaces into their own package Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Start renaming Storage methods Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the boltdb specific code from the Portainer storage code (example, the others need the same) Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extract bolt.Tx from datastore code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * minimise imports by putting moving the struct definition into the file that needs the Service imports Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extraction of boltdb.Tx Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the use of bucket.SetSequence Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * almost done - just endpoint.Synchonise :/ Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * so, endpoint.Synchonize looks hard, but i can't find where we use it, so 'delete first refactoring' Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix test compile errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * test compile fixes after rebase Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix a mis-remembering I had wrt deserialisation - last time i used AnyData - jsoniter's bindTo looks interesting for the same reason Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * set us up to make the connection an interface Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make the db connection a datastore interface, and separate out our datastore services from the bolt ones Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * rename methods to something less oltdb internals specific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * these errors are not boltdb secific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * start using the db-backend factory method too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * export boltdb raw in case we can't export from the service layer Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add a raw export from boltdb to yaml for broken db's, and an export services to yaml in backup Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add the version info by hand for now Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * actually, the export from services can be fully typed - its the import that needs to do more work Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * redo raw export, and make import capable of using it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add DockerHub Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * migration from anything older than v1.21.0 has been broken for quite a while, deleting the un-tested code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix go test ./... again Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * my goland wasn't setup to gofmt Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the two extremely dubious migration tests down into store, so they can use the test store code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * the migrator is now free of boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * reverse goland overzealous replcement of internal with boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more undo over-zealous goland internal->boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * yay, now bolt is only mentioned inside the api/database/ dir Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * and this might be the last of the boltdb references? Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add todo Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the store code into a separate module too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * don't need the fileService in boltdb anymore Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use IsErrObjectNotFound() Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use a string to select what database backend we use Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make isNew store an ephemeral bool that doesn't stay true after we've initialised it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the import.json wip to a separate file so its more obvious - we'll be using it for testing, emergency fixups, and in the next part of the store work, when we improve migrations and data model lifecycles Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * undo vscode formatting html Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Update missing helm user repo methods * feat(store): implement store tests EE-2112 (#6224) * add store tests * add some more tests * remove redundant comments * add webhook export * update webhooks * fix build issues after rebasing * move migratorparams * remove unneeded integer type conversions * disable the db import/export for now Co-authored-by: Richard Wei <54336863+WaysonWei@users.noreply.github.com> Co-authored-by: cong meng <mcpacino@gmail.com> Co-authored-by: Simon Meng <simon.meng@portainer.io> Co-authored-by: Marcelo Rydel <marcelorydel26@gmail.com> Co-authored-by: Hao Zhang <hao.zhang@portainer.io> Co-authored-by: sunportainer <93502624+sunportainer@users.noreply.github.com> Co-authored-by: sunportainer <ericsun@SG1.local> Co-authored-by: wheresolivia <78844659+wheresolivia@users.noreply.github.com> Co-authored-by: Chaim Lev-Ari <chiptus@users.noreply.github.com> Co-authored-by: Chao Geng <93526589+chaogeng77977@users.noreply.github.com> Co-authored-by: Dmitry Salakhov <to@dimasalakhov.com> Co-authored-by: Matt Hook <hookenz@gmail.com>
2021-12-15 02:26:09 +00:00
"github.com/portainer/portainer/api/dataservices"
"github.com/portainer/portainer/api/docker"
"github.com/portainer/portainer/api/http/security"
"github.com/portainer/portainer/api/internal/authorization"
"github.com/portainer/portainer/api/kubernetes/cli"
"github.com/portainer/portainer/api/scheduler"
"github.com/portainer/portainer/api/stacks"
)
var (
errStackAlreadyExists = errors.New("A stack already exists with this name")
errWebhookIDAlreadyExists = errors.New("A webhook ID already exists")
errStackNotExternal = errors.New("Not an external stack")
)
// Handler is the HTTP handler used to handle stack operations.
type Handler struct {
stackCreationMutex *sync.Mutex
stackDeletionMutex *sync.Mutex
requestBouncer *security.RequestBouncer
*mux.Router
DataStore dataservices.DataStore
DockerClientFactory *docker.ClientFactory
FileService portainer.FileService
GitService portainer.GitService
SwarmStackManager portainer.SwarmStackManager
ComposeStackManager portainer.ComposeStackManager
KubernetesDeployer portainer.KubernetesDeployer
KubernetesClientFactory *cli.ClientFactory
Scheduler *scheduler.Scheduler
StackDeployer stacks.StackDeployer
}
chore(store) EE-1981: Refactor/store/error checking, and other refactoring (#6173) * use the Store interface IsErrObjectNotFound() to avoid revealing internal errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * what happens when you extract the datastore interfaces into their own package Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Start renaming Storage methods Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the boltdb specific code from the Portainer storage code (example, the others need the same) Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extract bolt.Tx from datastore code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * minimise imports by putting moving the struct definition into the file that needs the Service imports Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extraction of boltdb.Tx Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the use of bucket.SetSequence Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * almost done - just endpoint.Synchonise :/ Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * so, endpoint.Synchonize looks hard, but i can't find where we use it, so 'delete first refactoring' Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix test compile errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * test compile fixes after rebase Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix a mis-remembering I had wrt deserialisation - last time i used AnyData - jsoniter's bindTo looks interesting for the same reason Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * set us up to make the connection an interface Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make the db connection a datastore interface, and separate out our datastore services from the bolt ones Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * rename methods to something less oltdb internals specific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * these errors are not boltdb secific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * start using the db-backend factory method too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * export boltdb raw in case we can't export from the service layer Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add a raw export from boltdb to yaml for broken db's, and an export services to yaml in backup Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add the version info by hand for now Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * actually, the export from services can be fully typed - its the import that needs to do more work Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * redo raw export, and make import capable of using it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add DockerHub Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * migration from anything older than v1.21.0 has been broken for quite a while, deleting the un-tested code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix go test ./... again Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * my goland wasn't setup to gofmt Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the two extremely dubious migration tests down into store, so they can use the test store code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * the migrator is now free of boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * reverse goland overzealous replcement of internal with boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more undo over-zealous goland internal->boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * yay, now bolt is only mentioned inside the api/database/ dir Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * and this might be the last of the boltdb references? Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add todo Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the store code into a separate module too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * don't need the fileService in boltdb anymore Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use IsErrObjectNotFound() Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use a string to select what database backend we use Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make isNew store an ephemeral bool that doesn't stay true after we've initialised it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the import.json wip to a separate file so its more obvious - we'll be using it for testing, emergency fixups, and in the next part of the store work, when we improve migrations and data model lifecycles Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * undo vscode formatting html Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix app templates symbol (#6221) * feat(webhook) EE-2125 send registry auth haeder when update swarms service via webhook (#6220) * feat(webhook) EE-2125 add some helpers to registry utils * feat(webhook) EE-2125 persist registryID when creating a webhook * feat(webhook) EE-2125 send registry auth header when executing a webhook * feat(webhook) EE-2125 send registryID to backend when creating a service with webhook * feat(webhook) EE-2125 use the initial registry ID to create webhook on editing service screen * feat(webhook) EE-2125 update webhook when update registry * feat(webhook) EE-2125 add endpoint of update webhook * feat(webhook) EE-2125 code cleanup * feat(webhook) EE-2125 fix a typo * feat(webhook) EE-2125 fix circle import issue with unit test Co-authored-by: Simon Meng <simon.meng@portainer.io> * fix(kubeconfig): show kubeconfig download button for non admin users [EE-2123] (#6204) Co-authored-by: Simon Meng <simon.meng@portainer.io> * fix data-cy for k8s cluster menu (#6226) LGTM * feat(stack): make stack created from app template editable EE-1941 (#6104) feat(stack): make stack from app template editable * fix(container):disable Duplicate/Edit button when the container is portainer (#6223) * fix/ee-1909/show-pull-image-error (#6195) Co-authored-by: sunportainer <ericsun@SG1.local> * feat(cy): add data-cy to helm install button (#6241) * feat(cy): add data-cy to add registry button (#6242) * refactor(app): convert root folder files to es6 (#4159) * refactor(app): duplicate constants as es6 exports (#4158) * fix(docker): provide workaround to save network name variable (#6080) * fix/EE-1862/unable-to-stop-or-remove-stack workaround for var without default value in yaml file * fix/EE-1862/unable-to-stop-or-remove-stack check yaml file * fixed func and var names * wrapper error and used bool for stringset * UT case for createNetworkEnvFile * UT case for %s=%s * powerful StringSet * wrapper error for extract network name * wrapper all the return err * store more env * put to env file * make default value None * feat: gzip static resources (#6258) * fix(ssl)//handle --sslcert and --sslkey ee-2106 (#6203) * fix/ee-2106/handle-sslcert-sslkey Co-authored-by: sunportainer <ericsun@SG1.local> * fix(server):support disable https only ee-2068 (#6232) * fix/ee-2068/disable-forcely-https * feat(store): implement store tests EE-2112 (#6224) * add store tests * add some more tests * Update missing helm user repo methods * remove redundant comments * add webhook export * update webhooks * use the Store interface IsErrObjectNotFound() to avoid revealing internal errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * what happens when you extract the datastore interfaces into their own package Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Start renaming Storage methods Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the boltdb specific code from the Portainer storage code (example, the others need the same) Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extract bolt.Tx from datastore code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * minimise imports by putting moving the struct definition into the file that needs the Service imports Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extraction of boltdb.Tx Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the use of bucket.SetSequence Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * almost done - just endpoint.Synchonise :/ Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * so, endpoint.Synchonize looks hard, but i can't find where we use it, so 'delete first refactoring' Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix test compile errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * test compile fixes after rebase Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix a mis-remembering I had wrt deserialisation - last time i used AnyData - jsoniter's bindTo looks interesting for the same reason Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * set us up to make the connection an interface Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make the db connection a datastore interface, and separate out our datastore services from the bolt ones Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * rename methods to something less oltdb internals specific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * these errors are not boltdb secific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * start using the db-backend factory method too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * export boltdb raw in case we can't export from the service layer Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add a raw export from boltdb to yaml for broken db's, and an export services to yaml in backup Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add the version info by hand for now Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * actually, the export from services can be fully typed - its the import that needs to do more work Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * redo raw export, and make import capable of using it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add DockerHub Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * migration from anything older than v1.21.0 has been broken for quite a while, deleting the un-tested code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix go test ./... again Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * my goland wasn't setup to gofmt Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the two extremely dubious migration tests down into store, so they can use the test store code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * the migrator is now free of boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * reverse goland overzealous replcement of internal with boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more undo over-zealous goland internal->boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * yay, now bolt is only mentioned inside the api/database/ dir Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * and this might be the last of the boltdb references? Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add todo Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the store code into a separate module too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * don't need the fileService in boltdb anymore Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use IsErrObjectNotFound() Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use a string to select what database backend we use Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make isNew store an ephemeral bool that doesn't stay true after we've initialised it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the import.json wip to a separate file so its more obvious - we'll be using it for testing, emergency fixups, and in the next part of the store work, when we improve migrations and data model lifecycles Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * undo vscode formatting html Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Update missing helm user repo methods * feat(store): implement store tests EE-2112 (#6224) * add store tests * add some more tests * remove redundant comments * add webhook export * update webhooks * fix build issues after rebasing * move migratorparams * remove unneeded integer type conversions * disable the db import/export for now Co-authored-by: Richard Wei <54336863+WaysonWei@users.noreply.github.com> Co-authored-by: cong meng <mcpacino@gmail.com> Co-authored-by: Simon Meng <simon.meng@portainer.io> Co-authored-by: Marcelo Rydel <marcelorydel26@gmail.com> Co-authored-by: Hao Zhang <hao.zhang@portainer.io> Co-authored-by: sunportainer <93502624+sunportainer@users.noreply.github.com> Co-authored-by: sunportainer <ericsun@SG1.local> Co-authored-by: wheresolivia <78844659+wheresolivia@users.noreply.github.com> Co-authored-by: Chaim Lev-Ari <chiptus@users.noreply.github.com> Co-authored-by: Chao Geng <93526589+chaogeng77977@users.noreply.github.com> Co-authored-by: Dmitry Salakhov <to@dimasalakhov.com> Co-authored-by: Matt Hook <hookenz@gmail.com>
2021-12-15 02:26:09 +00:00
func stackExistsError(name string) *httperror.HandlerError {
msg := fmt.Sprintf("A stack with the normalized name '%s' already exists", name)
err := errors.New(msg)
return &httperror.HandlerError{StatusCode: http.StatusConflict, Message: msg, Err: err}
}
// NewHandler creates a handler to manage stack operations.
func NewHandler(bouncer *security.RequestBouncer) *Handler {
h := &Handler{
Router: mux.NewRouter(),
stackCreationMutex: &sync.Mutex{},
stackDeletionMutex: &sync.Mutex{},
requestBouncer: bouncer,
}
h.Handle("/stacks",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackCreate))).Methods(http.MethodPost)
h.Handle("/stacks",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackList))).Methods(http.MethodGet)
h.Handle("/stacks/{id}",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackInspect))).Methods(http.MethodGet)
h.Handle("/stacks/{id}",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackDelete))).Methods(http.MethodDelete)
h.Handle("/stacks/{id}/associate",
bouncer.AdminAccess(httperror.LoggerHandler(h.stackAssociate))).Methods(http.MethodPut)
h.Handle("/stacks/{id}",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackUpdate))).Methods(http.MethodPut)
h.Handle("/stacks/{id}/git",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackUpdateGit))).Methods(http.MethodPost)
h.Handle("/stacks/{id}/git/redeploy",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackGitRedeploy))).Methods(http.MethodPut)
h.Handle("/stacks/{id}/file",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackFile))).Methods(http.MethodGet)
h.Handle("/stacks/{id}/migrate",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackMigrate))).Methods(http.MethodPost)
h.Handle("/stacks/{id}/start",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackStart))).Methods(http.MethodPost)
h.Handle("/stacks/{id}/stop",
bouncer.AuthenticatedAccess(httperror.LoggerHandler(h.stackStop))).Methods(http.MethodPost)
h.Handle("/stacks/webhooks/{webhookID}",
httperror.LoggerHandler(h.webhookInvoke)).Methods(http.MethodPost)
return h
}
feat(api): rewrite access control management in Docker (#3337) * feat(api): decorate Docker resource creation response with resource control * fix(api): fix a potential resource control conflict between stacks/volumes * feat(api): generate a default private resource control instead of admin only * fix(api): fix default RC value * fix(api): update RC authorizations check to support admin only flag * refactor(api): relocate access control related methods * fix(api): fix a potential conflict when fetching RC from database * refactor(api): refactor access control logic * refactor(api): remove the concept of DecoratedStack * feat(api): automatically remove RC when removing a Docker resource * refactor(api): update filter resource methods documentation * refactor(api): update proxy package structure * refactor(api): renamed proxy/misc package * feat(api): re-introduce ResourceControlDelete operation as admin restricted * refactor(api): relocate default endpoint authorizations * feat(api): migrate RBAC data * feat(app): ResourceControl management refactor * fix(api): fix access control issue on stack deletion and automatically delete RC * fix(api): fix stack filtering * fix(api): fix UpdateResourceControl operation checks * refactor(api): introduce a NewTransport builder method * refactor(api): inject endpoint in Docker transport * refactor(api): introduce Docker client into Docker transport * refactor(api): refactor http/proxy package * feat(api): inspect a Docker resource labels during access control validation * fix(api): only apply automatic resource control creation on success response * fix(api): fix stack access control check * fix(api): use StatusCreated instead of StatusOK for automatic resource control creation * fix(app): resource control fixes * fix(api): fix an issue preventing administrator to inspect a resource with a RC * refactor(api): remove useless error return * refactor(api): document DecorateStacks function * fix(api): fix invalid resource control type for container deletion * feat(api): support Docker system networks * feat(api): update Swagger docs * refactor(api): rename transport variable * refactor(api): rename transport variable * feat(networks): add system tag for system networks * feat(api): add support for resource control labels * feat(api): upgrade to DBVersion 22 * refactor(api): refactor access control management in Docker proxy * refactor(api): re-implement docker proxy taskListOperation * refactor(api): review parameters declaration * refactor(api): remove extra blank line * refactor(api): review method comments * fix(api): fix invalid ServerAddress property and review method visibility * feat(api): update error message * feat(api): update restrictedVolumeBrowserOperation method * refactor(api): refactor method parameters * refactor(api): minor refactor * refactor(api): change Azure transport visibility * refactor(api): update struct documentation * refactor(api): update struct documentation * feat(api): review restrictedResourceOperation method * refactor(api): remove unused authorization methods * feat(api): apply RBAC when enabled on stack operations * fix(api): fix invalid data migration procedure for DBVersion = 22 * fix(app): RC duplicate on private resource * feat(api): change Docker API version logic for libcompose/client factory * fix(api): update access denied error message to be Docker API compliant * fix(api): update volume browsing authorizations data migration * fix(api): fix an issue with access control in multi-node agent Swarm cluster
2019-11-12 23:41:42 +00:00
func (handler *Handler) userCanAccessStack(securityContext *security.RestrictedRequestContext, endpointID portainer.EndpointID, resourceControl *portainer.ResourceControl) (bool, error) {
user, err := handler.DataStore.User().User(securityContext.UserID)
if err != nil {
return false, err
feat(api): rewrite access control management in Docker (#3337) * feat(api): decorate Docker resource creation response with resource control * fix(api): fix a potential resource control conflict between stacks/volumes * feat(api): generate a default private resource control instead of admin only * fix(api): fix default RC value * fix(api): update RC authorizations check to support admin only flag * refactor(api): relocate access control related methods * fix(api): fix a potential conflict when fetching RC from database * refactor(api): refactor access control logic * refactor(api): remove the concept of DecoratedStack * feat(api): automatically remove RC when removing a Docker resource * refactor(api): update filter resource methods documentation * refactor(api): update proxy package structure * refactor(api): renamed proxy/misc package * feat(api): re-introduce ResourceControlDelete operation as admin restricted * refactor(api): relocate default endpoint authorizations * feat(api): migrate RBAC data * feat(app): ResourceControl management refactor * fix(api): fix access control issue on stack deletion and automatically delete RC * fix(api): fix stack filtering * fix(api): fix UpdateResourceControl operation checks * refactor(api): introduce a NewTransport builder method * refactor(api): inject endpoint in Docker transport * refactor(api): introduce Docker client into Docker transport * refactor(api): refactor http/proxy package * feat(api): inspect a Docker resource labels during access control validation * fix(api): only apply automatic resource control creation on success response * fix(api): fix stack access control check * fix(api): use StatusCreated instead of StatusOK for automatic resource control creation * fix(app): resource control fixes * fix(api): fix an issue preventing administrator to inspect a resource with a RC * refactor(api): remove useless error return * refactor(api): document DecorateStacks function * fix(api): fix invalid resource control type for container deletion * feat(api): support Docker system networks * feat(api): update Swagger docs * refactor(api): rename transport variable * refactor(api): rename transport variable * feat(networks): add system tag for system networks * feat(api): add support for resource control labels * feat(api): upgrade to DBVersion 22 * refactor(api): refactor access control management in Docker proxy * refactor(api): re-implement docker proxy taskListOperation * refactor(api): review parameters declaration * refactor(api): remove extra blank line * refactor(api): review method comments * fix(api): fix invalid ServerAddress property and review method visibility * feat(api): update error message * feat(api): update restrictedVolumeBrowserOperation method * refactor(api): refactor method parameters * refactor(api): minor refactor * refactor(api): change Azure transport visibility * refactor(api): update struct documentation * refactor(api): update struct documentation * feat(api): review restrictedResourceOperation method * refactor(api): remove unused authorization methods * feat(api): apply RBAC when enabled on stack operations * fix(api): fix invalid data migration procedure for DBVersion = 22 * fix(app): RC duplicate on private resource * feat(api): change Docker API version logic for libcompose/client factory * fix(api): update access denied error message to be Docker API compliant * fix(api): update volume browsing authorizations data migration * fix(api): fix an issue with access control in multi-node agent Swarm cluster
2019-11-12 23:41:42 +00:00
}
userTeamIDs := make([]portainer.TeamID, 0)
for _, membership := range securityContext.UserMemberships {
userTeamIDs = append(userTeamIDs, membership.TeamID)
}
if resourceControl != nil && authorization.UserCanAccessResource(securityContext.UserID, userTeamIDs, resourceControl) {
feat(api): rewrite access control management in Docker (#3337) * feat(api): decorate Docker resource creation response with resource control * fix(api): fix a potential resource control conflict between stacks/volumes * feat(api): generate a default private resource control instead of admin only * fix(api): fix default RC value * fix(api): update RC authorizations check to support admin only flag * refactor(api): relocate access control related methods * fix(api): fix a potential conflict when fetching RC from database * refactor(api): refactor access control logic * refactor(api): remove the concept of DecoratedStack * feat(api): automatically remove RC when removing a Docker resource * refactor(api): update filter resource methods documentation * refactor(api): update proxy package structure * refactor(api): renamed proxy/misc package * feat(api): re-introduce ResourceControlDelete operation as admin restricted * refactor(api): relocate default endpoint authorizations * feat(api): migrate RBAC data * feat(app): ResourceControl management refactor * fix(api): fix access control issue on stack deletion and automatically delete RC * fix(api): fix stack filtering * fix(api): fix UpdateResourceControl operation checks * refactor(api): introduce a NewTransport builder method * refactor(api): inject endpoint in Docker transport * refactor(api): introduce Docker client into Docker transport * refactor(api): refactor http/proxy package * feat(api): inspect a Docker resource labels during access control validation * fix(api): only apply automatic resource control creation on success response * fix(api): fix stack access control check * fix(api): use StatusCreated instead of StatusOK for automatic resource control creation * fix(app): resource control fixes * fix(api): fix an issue preventing administrator to inspect a resource with a RC * refactor(api): remove useless error return * refactor(api): document DecorateStacks function * fix(api): fix invalid resource control type for container deletion * feat(api): support Docker system networks * feat(api): update Swagger docs * refactor(api): rename transport variable * refactor(api): rename transport variable * feat(networks): add system tag for system networks * feat(api): add support for resource control labels * feat(api): upgrade to DBVersion 22 * refactor(api): refactor access control management in Docker proxy * refactor(api): re-implement docker proxy taskListOperation * refactor(api): review parameters declaration * refactor(api): remove extra blank line * refactor(api): review method comments * fix(api): fix invalid ServerAddress property and review method visibility * feat(api): update error message * feat(api): update restrictedVolumeBrowserOperation method * refactor(api): refactor method parameters * refactor(api): minor refactor * refactor(api): change Azure transport visibility * refactor(api): update struct documentation * refactor(api): update struct documentation * feat(api): review restrictedResourceOperation method * refactor(api): remove unused authorization methods * feat(api): apply RBAC when enabled on stack operations * fix(api): fix invalid data migration procedure for DBVersion = 22 * fix(app): RC duplicate on private resource * feat(api): change Docker API version logic for libcompose/client factory * fix(api): update access denied error message to be Docker API compliant * fix(api): update volume browsing authorizations data migration * fix(api): fix an issue with access control in multi-node agent Swarm cluster
2019-11-12 23:41:42 +00:00
return true, nil
}
return handler.userIsAdminOrEndpointAdmin(user, endpointID)
feat(api): rewrite access control management in Docker (#3337) * feat(api): decorate Docker resource creation response with resource control * fix(api): fix a potential resource control conflict between stacks/volumes * feat(api): generate a default private resource control instead of admin only * fix(api): fix default RC value * fix(api): update RC authorizations check to support admin only flag * refactor(api): relocate access control related methods * fix(api): fix a potential conflict when fetching RC from database * refactor(api): refactor access control logic * refactor(api): remove the concept of DecoratedStack * feat(api): automatically remove RC when removing a Docker resource * refactor(api): update filter resource methods documentation * refactor(api): update proxy package structure * refactor(api): renamed proxy/misc package * feat(api): re-introduce ResourceControlDelete operation as admin restricted * refactor(api): relocate default endpoint authorizations * feat(api): migrate RBAC data * feat(app): ResourceControl management refactor * fix(api): fix access control issue on stack deletion and automatically delete RC * fix(api): fix stack filtering * fix(api): fix UpdateResourceControl operation checks * refactor(api): introduce a NewTransport builder method * refactor(api): inject endpoint in Docker transport * refactor(api): introduce Docker client into Docker transport * refactor(api): refactor http/proxy package * feat(api): inspect a Docker resource labels during access control validation * fix(api): only apply automatic resource control creation on success response * fix(api): fix stack access control check * fix(api): use StatusCreated instead of StatusOK for automatic resource control creation * fix(app): resource control fixes * fix(api): fix an issue preventing administrator to inspect a resource with a RC * refactor(api): remove useless error return * refactor(api): document DecorateStacks function * fix(api): fix invalid resource control type for container deletion * feat(api): support Docker system networks * feat(api): update Swagger docs * refactor(api): rename transport variable * refactor(api): rename transport variable * feat(networks): add system tag for system networks * feat(api): add support for resource control labels * feat(api): upgrade to DBVersion 22 * refactor(api): refactor access control management in Docker proxy * refactor(api): re-implement docker proxy taskListOperation * refactor(api): review parameters declaration * refactor(api): remove extra blank line * refactor(api): review method comments * fix(api): fix invalid ServerAddress property and review method visibility * feat(api): update error message * feat(api): update restrictedVolumeBrowserOperation method * refactor(api): refactor method parameters * refactor(api): minor refactor * refactor(api): change Azure transport visibility * refactor(api): update struct documentation * refactor(api): update struct documentation * feat(api): review restrictedResourceOperation method * refactor(api): remove unused authorization methods * feat(api): apply RBAC when enabled on stack operations * fix(api): fix invalid data migration procedure for DBVersion = 22 * fix(app): RC duplicate on private resource * feat(api): change Docker API version logic for libcompose/client factory * fix(api): update access denied error message to be Docker API compliant * fix(api): update volume browsing authorizations data migration * fix(api): fix an issue with access control in multi-node agent Swarm cluster
2019-11-12 23:41:42 +00:00
}
func (handler *Handler) userIsAdmin(userID portainer.UserID) (bool, error) {
user, err := handler.DataStore.User().User(userID)
if err != nil {
return false, err
}
isAdmin := user.Role == portainer.AdministratorRole
return isAdmin, nil
}
func (handler *Handler) userIsAdminOrEndpointAdmin(user *portainer.User, endpointID portainer.EndpointID) (bool, error) {
isAdmin := user.Role == portainer.AdministratorRole
return isAdmin, nil
}
func (handler *Handler) userCanCreateStack(securityContext *security.RestrictedRequestContext, endpointID portainer.EndpointID) (bool, error) {
user, err := handler.DataStore.User().User(securityContext.UserID)
if err != nil {
return false, err
}
return handler.userIsAdminOrEndpointAdmin(user, endpointID)
}
feat(k8s): support git automated sync for k8s applications [EE-577] (#5548) * feat(stack): backport changes to CE EE-1189 * feat(stack): front end backport changes to CE EE-1199 (#5455) * feat(stack): front end backport changes to CE EE-1199 * fix k8s deploy logic * fixed web editor confirmation message typo. EE-1501 * fix(stack): fixed issue auth detail not remembered EE-1502 (#5459) * show status in buttons * removed onChangeRef function. * moved buttons in git form to its own component * removed unused variable. Co-authored-by: ArrisLee <arris_li@hotmail.com> * moved formvalue to kube app component * fix(stack): failed to pull and redeploy compose format k8s stack * fixed form value * fix(k8s): file content overridden when deployment failed with compose format EE-1548 * updated API response to get IsComposeFormat and show appropriate text. * feat(k8s): front end backport to CE * feat(kube): kube app auto update backend (#5547) * error message updates for different file type * not display creation source for external application * added confirmation modal to advanced app created by web editor * stop showing confirmation modal when updating application * disable rollback button when application type is not applicatiom form * only update file after deployment succeded * Revert "only update file after deployment succeded" This reverts commit b94bd2e96f27ca337e9d2fa698da2ab264d9ac86. * fix(k8s): file content overridden when deployment failed with compose format EE-1556 * added analytics-on directive to pull and redeploy button * fix(kube): don't valide resource control access for kube (#5568) * added missing question mark to k8s confirmation modal * fixed webhook format issue * added question marks to k8s app confirmation modal * added space in additional file list. * ignoring error on deletion * fix(k8s): Git authentication info not persisted * added RepositoryMechanismTypes constant * updated analytics functions * covert RepositoryMechanism to constant * fixed typo * removed unused function. * post tech review updates * fixed save settings n redeploy button * refact kub deploy logic * Revert "refact kub deploy logic" This reverts commit cbfdd58eceed438b148c836ed09004dd270d0aa6. * feat(k8s): utilize user token for k8s auto update EE-1594 * feat(k8s): persist kub stack name EE-1630 * feat(k8s): support delete kub stack * fix(app): updated logic to delete stack for different kind apps. (#5648) * fix(app): updated logic to delete stack for different kind apps. * renamed variable * fix import * added StackName field. * fixed stack id not found issue. * fix(k8s): fixed qusetion mark alignment issue in PAT field. (#5611) * fix(k8s): fixed qusetion mark alignment issue in PAT field. * moved inline css to file. * fix(git-form: made auth input text full width * add ignore deleted arg * tech review updates * typo fix * fix(k8s): added console error when deleting k8s service. * fix(console): added no-console config * fix(deploy): added missing service. * fix: use stack editor as an owner when exists (#5678) * fix: tempalte/content based stacks edit/delete * fix(stack): remove stack when no app. (#5769) * fix(stack): remove stack when no app. * support compose format in delete Co-authored-by: ArrisLee <arris_li@hotmail.com> Co-authored-by: Hui <arris_li@hotmail.com> Co-authored-by: fhanportainer <79428273+fhanportainer@users.noreply.github.com> Co-authored-by: Felix Han <felix.han@portainer.io>
2021-09-29 23:58:10 +00:00
func (handler *Handler) checkUniqueStackName(endpoint *portainer.Endpoint, name string, stackID portainer.StackID) (bool, error) {
stacks, err := handler.DataStore.Stack().Stacks()
if err != nil {
return false, err
}
for _, stack := range stacks {
if strings.EqualFold(stack.Name, name) && (stackID == 0 || stackID != stack.ID) && stack.EndpointID == endpoint.ID {
return false, nil
}
}
feat(k8s): support git automated sync for k8s applications [EE-577] (#5548) * feat(stack): backport changes to CE EE-1189 * feat(stack): front end backport changes to CE EE-1199 (#5455) * feat(stack): front end backport changes to CE EE-1199 * fix k8s deploy logic * fixed web editor confirmation message typo. EE-1501 * fix(stack): fixed issue auth detail not remembered EE-1502 (#5459) * show status in buttons * removed onChangeRef function. * moved buttons in git form to its own component * removed unused variable. Co-authored-by: ArrisLee <arris_li@hotmail.com> * moved formvalue to kube app component * fix(stack): failed to pull and redeploy compose format k8s stack * fixed form value * fix(k8s): file content overridden when deployment failed with compose format EE-1548 * updated API response to get IsComposeFormat and show appropriate text. * feat(k8s): front end backport to CE * feat(kube): kube app auto update backend (#5547) * error message updates for different file type * not display creation source for external application * added confirmation modal to advanced app created by web editor * stop showing confirmation modal when updating application * disable rollback button when application type is not applicatiom form * only update file after deployment succeded * Revert "only update file after deployment succeded" This reverts commit b94bd2e96f27ca337e9d2fa698da2ab264d9ac86. * fix(k8s): file content overridden when deployment failed with compose format EE-1556 * added analytics-on directive to pull and redeploy button * fix(kube): don't valide resource control access for kube (#5568) * added missing question mark to k8s confirmation modal * fixed webhook format issue * added question marks to k8s app confirmation modal * added space in additional file list. * ignoring error on deletion * fix(k8s): Git authentication info not persisted * added RepositoryMechanismTypes constant * updated analytics functions * covert RepositoryMechanism to constant * fixed typo * removed unused function. * post tech review updates * fixed save settings n redeploy button * refact kub deploy logic * Revert "refact kub deploy logic" This reverts commit cbfdd58eceed438b148c836ed09004dd270d0aa6. * feat(k8s): utilize user token for k8s auto update EE-1594 * feat(k8s): persist kub stack name EE-1630 * feat(k8s): support delete kub stack * fix(app): updated logic to delete stack for different kind apps. (#5648) * fix(app): updated logic to delete stack for different kind apps. * renamed variable * fix import * added StackName field. * fixed stack id not found issue. * fix(k8s): fixed qusetion mark alignment issue in PAT field. (#5611) * fix(k8s): fixed qusetion mark alignment issue in PAT field. * moved inline css to file. * fix(git-form: made auth input text full width * add ignore deleted arg * tech review updates * typo fix * fix(k8s): added console error when deleting k8s service. * fix(console): added no-console config * fix(deploy): added missing service. * fix: use stack editor as an owner when exists (#5678) * fix: tempalte/content based stacks edit/delete * fix(stack): remove stack when no app. (#5769) * fix(stack): remove stack when no app. * support compose format in delete Co-authored-by: ArrisLee <arris_li@hotmail.com> Co-authored-by: Hui <arris_li@hotmail.com> Co-authored-by: fhanportainer <79428273+fhanportainer@users.noreply.github.com> Co-authored-by: Felix Han <felix.han@portainer.io>
2021-09-29 23:58:10 +00:00
return true, nil
}
func (handler *Handler) checkUniqueStackNameInKubernetes(endpoint *portainer.Endpoint, name string, stackID portainer.StackID, namespace string) (bool, error) {
isUniqueStackName, err := handler.checkUniqueStackName(endpoint, name, stackID)
if err != nil {
return false, err
}
if !isUniqueStackName {
// Check if this stack name is really used in the kubernetes.
// Because the stack with this name could be removed via kubectl cli outside and the datastore does not be informed of this action.
if namespace == "" {
namespace = "default"
}
kubeCli, err := handler.KubernetesClientFactory.GetKubeClient(endpoint)
if err != nil {
return false, err
}
isUniqueStackName, err = kubeCli.HasStackName(namespace, name)
if err != nil {
return false, err
}
}
return isUniqueStackName, nil
}
feat(k8s): support git automated sync for k8s applications [EE-577] (#5548) * feat(stack): backport changes to CE EE-1189 * feat(stack): front end backport changes to CE EE-1199 (#5455) * feat(stack): front end backport changes to CE EE-1199 * fix k8s deploy logic * fixed web editor confirmation message typo. EE-1501 * fix(stack): fixed issue auth detail not remembered EE-1502 (#5459) * show status in buttons * removed onChangeRef function. * moved buttons in git form to its own component * removed unused variable. Co-authored-by: ArrisLee <arris_li@hotmail.com> * moved formvalue to kube app component * fix(stack): failed to pull and redeploy compose format k8s stack * fixed form value * fix(k8s): file content overridden when deployment failed with compose format EE-1548 * updated API response to get IsComposeFormat and show appropriate text. * feat(k8s): front end backport to CE * feat(kube): kube app auto update backend (#5547) * error message updates for different file type * not display creation source for external application * added confirmation modal to advanced app created by web editor * stop showing confirmation modal when updating application * disable rollback button when application type is not applicatiom form * only update file after deployment succeded * Revert "only update file after deployment succeded" This reverts commit b94bd2e96f27ca337e9d2fa698da2ab264d9ac86. * fix(k8s): file content overridden when deployment failed with compose format EE-1556 * added analytics-on directive to pull and redeploy button * fix(kube): don't valide resource control access for kube (#5568) * added missing question mark to k8s confirmation modal * fixed webhook format issue * added question marks to k8s app confirmation modal * added space in additional file list. * ignoring error on deletion * fix(k8s): Git authentication info not persisted * added RepositoryMechanismTypes constant * updated analytics functions * covert RepositoryMechanism to constant * fixed typo * removed unused function. * post tech review updates * fixed save settings n redeploy button * refact kub deploy logic * Revert "refact kub deploy logic" This reverts commit cbfdd58eceed438b148c836ed09004dd270d0aa6. * feat(k8s): utilize user token for k8s auto update EE-1594 * feat(k8s): persist kub stack name EE-1630 * feat(k8s): support delete kub stack * fix(app): updated logic to delete stack for different kind apps. (#5648) * fix(app): updated logic to delete stack for different kind apps. * renamed variable * fix import * added StackName field. * fixed stack id not found issue. * fix(k8s): fixed qusetion mark alignment issue in PAT field. (#5611) * fix(k8s): fixed qusetion mark alignment issue in PAT field. * moved inline css to file. * fix(git-form: made auth input text full width * add ignore deleted arg * tech review updates * typo fix * fix(k8s): added console error when deleting k8s service. * fix(console): added no-console config * fix(deploy): added missing service. * fix: use stack editor as an owner when exists (#5678) * fix: tempalte/content based stacks edit/delete * fix(stack): remove stack when no app. (#5769) * fix(stack): remove stack when no app. * support compose format in delete Co-authored-by: ArrisLee <arris_li@hotmail.com> Co-authored-by: Hui <arris_li@hotmail.com> Co-authored-by: fhanportainer <79428273+fhanportainer@users.noreply.github.com> Co-authored-by: Felix Han <felix.han@portainer.io>
2021-09-29 23:58:10 +00:00
func (handler *Handler) checkUniqueStackNameInDocker(endpoint *portainer.Endpoint, name string, stackID portainer.StackID, swarmMode bool) (bool, error) {
isUniqueStackName, err := handler.checkUniqueStackName(endpoint, name, stackID)
if err != nil {
return false, err
}
feat(intel): Enable OpenAMT and FDO capabilities (#6212) * feat(openamt): add AMT Devices information in Environments view [INT-8] (#6169) * feat(openamt): add AMT Devices Ouf of Band Managamenet actions [INT-9] (#6171) * feat(openamt): add AMT Devices KVM Connection [INT-10] (#6179) * feat(openamt): Enhance the Environments MX to activate OpenAMT on compatible environments [INT-7] (#6196) * feat(openamt): Enable KVM by default [INT-25] (#6228) * feat(fdo): implement the FDO configuration settings INT-19 (#6238) feat(fdo): implement the FDO configuration settings INT-19 * feat(fdo): implement Owner client INT-17 (#6231) feat(fdo): implement Owner client INT-17 * feat(openamt): hide wireless config in OpenAMT form (#6250) * feat(openamt): Increase OpenAMT timeouts [INT-30] (#6253) * feat(openamt): Disable the ability to use KVM and OOB actions on a MPS disconnected device [INT-36] (#6254) * feat(fdo): add import device UI [INT-20] (#6240) feat(fdo): add import device UI INT-20 * refactor(fdo): fix develop merge issues * feat(openamt): Do not fetch OpenAMT details for an unassociated Edge endpoint (#6273) * fix(intel): Fix switches params (#6282) * feat(openamt): preload existing AMT settings (#6283) * feat(openamt): Better UI/UX for AMT activation loading [INT-39] (#6290) * feat(openamt): Remove wireless config related code [INT-41] (#6291) * yarn install * feat(openamt): change kvm redirection for pop up, always enable features [INT-37] (#6292) * feat(openamt): change kvm redirection for pop up, always enable features [INT-37] (#6293) * feat(openmt): use .ts services with axios for OpenAMT (#6312) * Minor code cleanup. * fix(fdo): move the FDO client code to the hostmanagement folder INT-44 (#6345) * refactor(intel): Add Edge Compute Settings view (#6351) * feat(fdo): add FDO profiles INT-22 (#6363) feat(fdo): add FDO profiles INT-22 * fix(fdo): fix incorrect profile URL INT-45 (#6377) * fixed husky version * fix go.mod with go mod tidy * feat(edge): migrate OpenAMT devices views to Edge Devices [EE-2322] (#6373) * feat(intel): OpenAMT UI/UX adjustments (#6394) * only allow edge agent as edge device * show all edge agent environments on Edge Devices view * feat(fdo): add the ability to import multiple ownership vouchers at once EE-2324 (#6395) * fix(edge): settings edge compute alert (#6402) * remove pagination, add useMemo for devices result array (#6409) * feat(edge): minor Edge Devices (AMT) UI fixes (#6410) * chore(eslint): fix versions * chore(app): reformat codebase * change add edge agent modal behaviour, fix yarn.lock * fix use pagination * remove extractedTranslations folder * feat(edge): add FDO Profiles Datatable [EE-2406] (#6415) * feat(edge): add KVM workaround tooltip (#6441) * feat(edge): Add default FDO profile (#6450) * feat(edge): add settings to disable trust on first connect and enforce Edge ID INT-1 EE-2410 (#6429) Co-authored-by: andres-portainer <91705312+andres-portainer@users.noreply.github.com> Co-authored-by: Anthony Lapenna <anthony.lapenna@portainer.io> Co-authored-by: andres-portainer <andres-portainer@users.noreply.github.com> Co-authored-by: Chaim Lev-Ari <chiptus@gmail.com>
2022-01-23 19:48:04 +00:00
dockerClient, err := handler.DockerClientFactory.CreateClient(endpoint, "", nil)
if err != nil {
return false, err
}
defer dockerClient.Close()
if swarmMode {
services, err := dockerClient.ServiceList(context.Background(), types.ServiceListOptions{})
if err != nil {
return false, err
}
for _, service := range services {
serviceNS, ok := service.Spec.Labels["com.docker.stack.namespace"]
if ok && serviceNS == name {
return false, nil
}
}
}
containers, err := dockerClient.ContainerList(context.Background(), types.ContainerListOptions{All: true})
if err != nil {
return false, err
}
for _, container := range containers {
containerNS, ok := container.Labels["com.docker.compose.project"]
if ok && containerNS == name {
return false, nil
}
}
feat(k8s): support git automated sync for k8s applications [EE-577] (#5548) * feat(stack): backport changes to CE EE-1189 * feat(stack): front end backport changes to CE EE-1199 (#5455) * feat(stack): front end backport changes to CE EE-1199 * fix k8s deploy logic * fixed web editor confirmation message typo. EE-1501 * fix(stack): fixed issue auth detail not remembered EE-1502 (#5459) * show status in buttons * removed onChangeRef function. * moved buttons in git form to its own component * removed unused variable. Co-authored-by: ArrisLee <arris_li@hotmail.com> * moved formvalue to kube app component * fix(stack): failed to pull and redeploy compose format k8s stack * fixed form value * fix(k8s): file content overridden when deployment failed with compose format EE-1548 * updated API response to get IsComposeFormat and show appropriate text. * feat(k8s): front end backport to CE * feat(kube): kube app auto update backend (#5547) * error message updates for different file type * not display creation source for external application * added confirmation modal to advanced app created by web editor * stop showing confirmation modal when updating application * disable rollback button when application type is not applicatiom form * only update file after deployment succeded * Revert "only update file after deployment succeded" This reverts commit b94bd2e96f27ca337e9d2fa698da2ab264d9ac86. * fix(k8s): file content overridden when deployment failed with compose format EE-1556 * added analytics-on directive to pull and redeploy button * fix(kube): don't valide resource control access for kube (#5568) * added missing question mark to k8s confirmation modal * fixed webhook format issue * added question marks to k8s app confirmation modal * added space in additional file list. * ignoring error on deletion * fix(k8s): Git authentication info not persisted * added RepositoryMechanismTypes constant * updated analytics functions * covert RepositoryMechanism to constant * fixed typo * removed unused function. * post tech review updates * fixed save settings n redeploy button * refact kub deploy logic * Revert "refact kub deploy logic" This reverts commit cbfdd58eceed438b148c836ed09004dd270d0aa6. * feat(k8s): utilize user token for k8s auto update EE-1594 * feat(k8s): persist kub stack name EE-1630 * feat(k8s): support delete kub stack * fix(app): updated logic to delete stack for different kind apps. (#5648) * fix(app): updated logic to delete stack for different kind apps. * renamed variable * fix import * added StackName field. * fixed stack id not found issue. * fix(k8s): fixed qusetion mark alignment issue in PAT field. (#5611) * fix(k8s): fixed qusetion mark alignment issue in PAT field. * moved inline css to file. * fix(git-form: made auth input text full width * add ignore deleted arg * tech review updates * typo fix * fix(k8s): added console error when deleting k8s service. * fix(console): added no-console config * fix(deploy): added missing service. * fix: use stack editor as an owner when exists (#5678) * fix: tempalte/content based stacks edit/delete * fix(stack): remove stack when no app. (#5769) * fix(stack): remove stack when no app. * support compose format in delete Co-authored-by: ArrisLee <arris_li@hotmail.com> Co-authored-by: Hui <arris_li@hotmail.com> Co-authored-by: fhanportainer <79428273+fhanportainer@users.noreply.github.com> Co-authored-by: Felix Han <felix.han@portainer.io>
2021-09-29 23:58:10 +00:00
return isUniqueStackName, nil
}
func (handler *Handler) checkUniqueWebhookID(webhookID string) (bool, error) {
_, err := handler.DataStore.Stack().StackByWebhookID(webhookID)
chore(store) EE-1981: Refactor/store/error checking, and other refactoring (#6173) * use the Store interface IsErrObjectNotFound() to avoid revealing internal errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * what happens when you extract the datastore interfaces into their own package Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Start renaming Storage methods Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the boltdb specific code from the Portainer storage code (example, the others need the same) Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extract bolt.Tx from datastore code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * minimise imports by putting moving the struct definition into the file that needs the Service imports Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extraction of boltdb.Tx Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the use of bucket.SetSequence Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * almost done - just endpoint.Synchonise :/ Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * so, endpoint.Synchonize looks hard, but i can't find where we use it, so 'delete first refactoring' Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix test compile errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * test compile fixes after rebase Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix a mis-remembering I had wrt deserialisation - last time i used AnyData - jsoniter's bindTo looks interesting for the same reason Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * set us up to make the connection an interface Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make the db connection a datastore interface, and separate out our datastore services from the bolt ones Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * rename methods to something less oltdb internals specific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * these errors are not boltdb secific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * start using the db-backend factory method too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * export boltdb raw in case we can't export from the service layer Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add a raw export from boltdb to yaml for broken db's, and an export services to yaml in backup Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add the version info by hand for now Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * actually, the export from services can be fully typed - its the import that needs to do more work Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * redo raw export, and make import capable of using it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add DockerHub Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * migration from anything older than v1.21.0 has been broken for quite a while, deleting the un-tested code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix go test ./... again Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * my goland wasn't setup to gofmt Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the two extremely dubious migration tests down into store, so they can use the test store code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * the migrator is now free of boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * reverse goland overzealous replcement of internal with boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more undo over-zealous goland internal->boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * yay, now bolt is only mentioned inside the api/database/ dir Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * and this might be the last of the boltdb references? Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add todo Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the store code into a separate module too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * don't need the fileService in boltdb anymore Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use IsErrObjectNotFound() Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use a string to select what database backend we use Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make isNew store an ephemeral bool that doesn't stay true after we've initialised it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the import.json wip to a separate file so its more obvious - we'll be using it for testing, emergency fixups, and in the next part of the store work, when we improve migrations and data model lifecycles Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * undo vscode formatting html Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix app templates symbol (#6221) * feat(webhook) EE-2125 send registry auth haeder when update swarms service via webhook (#6220) * feat(webhook) EE-2125 add some helpers to registry utils * feat(webhook) EE-2125 persist registryID when creating a webhook * feat(webhook) EE-2125 send registry auth header when executing a webhook * feat(webhook) EE-2125 send registryID to backend when creating a service with webhook * feat(webhook) EE-2125 use the initial registry ID to create webhook on editing service screen * feat(webhook) EE-2125 update webhook when update registry * feat(webhook) EE-2125 add endpoint of update webhook * feat(webhook) EE-2125 code cleanup * feat(webhook) EE-2125 fix a typo * feat(webhook) EE-2125 fix circle import issue with unit test Co-authored-by: Simon Meng <simon.meng@portainer.io> * fix(kubeconfig): show kubeconfig download button for non admin users [EE-2123] (#6204) Co-authored-by: Simon Meng <simon.meng@portainer.io> * fix data-cy for k8s cluster menu (#6226) LGTM * feat(stack): make stack created from app template editable EE-1941 (#6104) feat(stack): make stack from app template editable * fix(container):disable Duplicate/Edit button when the container is portainer (#6223) * fix/ee-1909/show-pull-image-error (#6195) Co-authored-by: sunportainer <ericsun@SG1.local> * feat(cy): add data-cy to helm install button (#6241) * feat(cy): add data-cy to add registry button (#6242) * refactor(app): convert root folder files to es6 (#4159) * refactor(app): duplicate constants as es6 exports (#4158) * fix(docker): provide workaround to save network name variable (#6080) * fix/EE-1862/unable-to-stop-or-remove-stack workaround for var without default value in yaml file * fix/EE-1862/unable-to-stop-or-remove-stack check yaml file * fixed func and var names * wrapper error and used bool for stringset * UT case for createNetworkEnvFile * UT case for %s=%s * powerful StringSet * wrapper error for extract network name * wrapper all the return err * store more env * put to env file * make default value None * feat: gzip static resources (#6258) * fix(ssl)//handle --sslcert and --sslkey ee-2106 (#6203) * fix/ee-2106/handle-sslcert-sslkey Co-authored-by: sunportainer <ericsun@SG1.local> * fix(server):support disable https only ee-2068 (#6232) * fix/ee-2068/disable-forcely-https * feat(store): implement store tests EE-2112 (#6224) * add store tests * add some more tests * Update missing helm user repo methods * remove redundant comments * add webhook export * update webhooks * use the Store interface IsErrObjectNotFound() to avoid revealing internal errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * what happens when you extract the datastore interfaces into their own package Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Start renaming Storage methods Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the boltdb specific code from the Portainer storage code (example, the others need the same) Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extract bolt.Tx from datastore code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * minimise imports by putting moving the struct definition into the file that needs the Service imports Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more extraction of boltdb.Tx Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the use of bucket.SetSequence Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * almost done - just endpoint.Synchonise :/ Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * so, endpoint.Synchonize looks hard, but i can't find where we use it, so 'delete first refactoring' Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix test compile errors Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * test compile fixes after rebase Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix a mis-remembering I had wrt deserialisation - last time i used AnyData - jsoniter's bindTo looks interesting for the same reason Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * set us up to make the connection an interface Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make the db connection a datastore interface, and separate out our datastore services from the bolt ones Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * rename methods to something less oltdb internals specific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * these errors are not boltdb secific Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * start using the db-backend factory method too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * export boltdb raw in case we can't export from the service layer Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add a raw export from boltdb to yaml for broken db's, and an export services to yaml in backup Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add the version info by hand for now Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * actually, the export from services can be fully typed - its the import that needs to do more work Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * redo raw export, and make import capable of using it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add DockerHub Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * migration from anything older than v1.21.0 has been broken for quite a while, deleting the un-tested code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * fix go test ./... again Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * my goland wasn't setup to gofmt Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the two extremely dubious migration tests down into store, so they can use the test store code Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * the migrator is now free of boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * reverse goland overzealous replcement of internal with boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * more undo over-zealous goland internal->boltdb Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * yay, now bolt is only mentioned inside the api/database/ dir Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * and this might be the last of the boltdb references? Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * add todo Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * extract the store code into a separate module too Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * don't need the fileService in boltdb anymore Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use IsErrObjectNotFound() Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * use a string to select what database backend we use Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * make isNew store an ephemeral bool that doesn't stay true after we've initialised it Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * move the import.json wip to a separate file so its more obvious - we'll be using it for testing, emergency fixups, and in the next part of the store work, when we improve migrations and data model lifecycles Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * undo vscode formatting html Signed-off-by: Sven Dowideit <sven.dowideit@portainer.io> * Update missing helm user repo methods * feat(store): implement store tests EE-2112 (#6224) * add store tests * add some more tests * remove redundant comments * add webhook export * update webhooks * fix build issues after rebasing * move migratorparams * remove unneeded integer type conversions * disable the db import/export for now Co-authored-by: Richard Wei <54336863+WaysonWei@users.noreply.github.com> Co-authored-by: cong meng <mcpacino@gmail.com> Co-authored-by: Simon Meng <simon.meng@portainer.io> Co-authored-by: Marcelo Rydel <marcelorydel26@gmail.com> Co-authored-by: Hao Zhang <hao.zhang@portainer.io> Co-authored-by: sunportainer <93502624+sunportainer@users.noreply.github.com> Co-authored-by: sunportainer <ericsun@SG1.local> Co-authored-by: wheresolivia <78844659+wheresolivia@users.noreply.github.com> Co-authored-by: Chaim Lev-Ari <chiptus@users.noreply.github.com> Co-authored-by: Chao Geng <93526589+chaogeng77977@users.noreply.github.com> Co-authored-by: Dmitry Salakhov <to@dimasalakhov.com> Co-authored-by: Matt Hook <hookenz@gmail.com>
2021-12-15 02:26:09 +00:00
if handler.DataStore.IsErrObjectNotFound(err) {
return true, nil
}
return false, err
}
func (handler *Handler) clone(projectPath, repositoryURL, refName string, auth bool, username, password string) error {
if !auth {
username = ""
password = ""
}
err := handler.GitService.CloneRepository(projectPath, repositoryURL, refName, username, password)
if err != nil {
return fmt.Errorf("unable to clone git repository: %w", err)
}
return nil
}
func (handler *Handler) latestCommitID(repositoryURL, refName string, auth bool, username, password string) (string, error) {
if !auth {
username = ""
password = ""
}
return handler.GitService.LatestCommitID(repositoryURL, refName, username, password)
}