fix: fix null pointer on external provisioner daemons with daily_cost (#9401)

* fix: fix null pointer on external provisioner daemons with daily_cost

Signed-off-by: Spike Curtis <spike@coder.com>

* Add logging for debounce and job acquire

Signed-off-by: Spike Curtis <spike@coder.com>

* Return error instead of panic

Signed-off-by: Spike Curtis <spike@coder.com>

* remove debounce on external provisioners to fix test flakes

Signed-off-by: Spike Curtis <spike@coder.com>

---------

Signed-off-by: Spike Curtis <spike@coder.com>
This commit is contained in:
Spike Curtis
2023-08-30 10:48:35 +00:00
committed by GitHub
parent a415395e9e
commit 90acf998bf
6 changed files with 664 additions and 350 deletions
+28 -17
View File
@@ -11,6 +11,7 @@ import (
"net"
"net/http"
"strings"
"time"
"github.com/google/uuid"
"github.com/hashicorp/yamux"
@@ -243,23 +244,33 @@ func (api *API) provisionerDaemonServe(rw http.ResponseWriter, r *http.Request)
return
}
mux := drpcmux.New()
err = proto.DRPCRegisterProvisionerDaemon(mux, &provisionerdserver.Server{
AccessURL: api.AccessURL,
GitAuthConfigs: api.GitAuthConfigs,
OIDCConfig: api.OIDCConfig,
ID: daemon.ID,
Database: api.Database,
Pubsub: api.Pubsub,
Provisioners: daemon.Provisioners,
Telemetry: api.Telemetry,
Auditor: &api.AGPL.Auditor,
TemplateScheduleStore: api.AGPL.TemplateScheduleStore,
UserQuietHoursScheduleStore: api.AGPL.UserQuietHoursScheduleStore,
Logger: api.Logger.Named(fmt.Sprintf("provisionerd-%s", daemon.Name)),
Tags: rawTags,
Tracer: trace.NewNoopTracerProvider().Tracer("noop"),
DeploymentValues: api.DeploymentValues,
})
srv, err := provisionerdserver.NewServer(
api.AccessURL,
daemon.ID,
api.Logger.Named(fmt.Sprintf("provisionerd-%s", daemon.Name)),
daemon.Provisioners,
rawTags,
api.Database,
api.Pubsub,
api.Telemetry,
trace.NewNoopTracerProvider().Tracer("noop"),
&api.AGPL.QuotaCommitter,
&api.AGPL.Auditor,
api.AGPL.TemplateScheduleStore,
api.AGPL.UserQuietHoursScheduleStore,
api.DeploymentValues,
// TODO(spikecurtis) - fix debounce to not cause flaky tests.
time.Duration(0),
provisionerdserver.Options{
GitAuthConfigs: api.GitAuthConfigs,
OIDCConfig: api.OIDCConfig,
},
)
if err != nil {
_ = conn.Close(websocket.StatusInternalError, httpapi.WebsocketCloseSprintf("create provisioner daemon server: %s", err))
return
}
err = proto.DRPCRegisterProvisionerDaemon(mux, srv)
if err != nil {
_ = conn.Close(websocket.StatusInternalError, httpapi.WebsocketCloseSprintf("drpc register provisioner daemon: %s", err))
return
@@ -9,13 +9,20 @@ import (
"github.com/google/uuid"
"github.com/stretchr/testify/require"
"cdr.dev/slog"
"cdr.dev/slog/sloggers/slogtest"
"github.com/coder/coder/v2/coderd/coderdtest"
"github.com/coder/coder/v2/coderd/database"
"github.com/coder/coder/v2/coderd/provisionerdserver"
"github.com/coder/coder/v2/coderd/rbac"
"github.com/coder/coder/v2/coderd/util/ptr"
"github.com/coder/coder/v2/codersdk"
"github.com/coder/coder/v2/enterprise/coderd/coderdenttest"
"github.com/coder/coder/v2/enterprise/coderd/license"
"github.com/coder/coder/v2/provisioner/echo"
"github.com/coder/coder/v2/provisionerd"
provisionerdproto "github.com/coder/coder/v2/provisionerd/proto"
"github.com/coder/coder/v2/provisionersdk"
"github.com/coder/coder/v2/provisionersdk/proto"
"github.com/coder/coder/v2/testutil"
)
@@ -212,6 +219,107 @@ func TestProvisionerDaemonServe(t *testing.T) {
require.Len(t, daemons, 1)
})
t.Run("PSK_daily_cost", func(t *testing.T) {
t.Parallel()
client, user := coderdenttest.New(t, &coderdenttest.Options{
UserWorkspaceQuota: 10,
LicenseOptions: &coderdenttest.LicenseOptions{
Features: license.Features{
codersdk.FeatureExternalProvisionerDaemons: 1,
codersdk.FeatureTemplateRBAC: 1,
},
},
ProvisionerDaemonPSK: "provisionersftw",
})
logger := slogtest.Make(t, nil).Leveled(slog.LevelDebug)
ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
defer cancel()
terraformClient, terraformServer := provisionersdk.MemTransportPipe()
go func() {
<-ctx.Done()
_ = terraformClient.Close()
_ = terraformServer.Close()
}()
tempDir := t.TempDir()
errCh := make(chan error)
go func() {
err := echo.Serve(ctx, &provisionersdk.ServeOptions{
Listener: terraformServer,
Logger: logger.Named("echo"),
WorkDirectory: tempDir,
})
errCh <- err
}()
provisioners := provisionerd.Provisioners{
string(database.ProvisionerTypeEcho): proto.NewDRPCProvisionerClient(terraformClient),
}
another := codersdk.New(client.URL)
pd := provisionerd.New(func(ctx context.Context) (provisionerdproto.DRPCProvisionerDaemonClient, error) {
return another.ServeProvisionerDaemon(ctx, codersdk.ServeProvisionerDaemonRequest{
Organization: user.OrganizationID,
Provisioners: []codersdk.ProvisionerType{
codersdk.ProvisionerTypeEcho,
},
Tags: map[string]string{
provisionerdserver.TagScope: provisionerdserver.ScopeOrganization,
},
PreSharedKey: "provisionersftw",
})
}, &provisionerd.Options{
Logger: logger.Named("provisionerd"),
Provisioners: provisioners,
})
defer pd.Close()
// Patch the 'Everyone' group to give the user quota to build their workspace.
_, err := client.PatchGroup(ctx, user.OrganizationID, codersdk.PatchGroupRequest{
QuotaAllowance: ptr.Ref(1),
})
require.NoError(t, err)
authToken := uuid.NewString()
version := coderdtest.CreateTemplateVersion(t, client, user.OrganizationID, &echo.Responses{
Parse: echo.ParseComplete,
ProvisionApply: []*proto.Response{{
Type: &proto.Response_Apply{
Apply: &proto.ApplyComplete{
Resources: []*proto.Resource{{
Name: "example",
Type: "aws_instance",
DailyCost: 1,
Agents: []*proto.Agent{{
Id: uuid.NewString(),
Name: "example",
Auth: &proto.Agent_Token{
Token: authToken,
},
}},
}},
},
},
}},
})
coderdtest.AwaitTemplateVersionJob(t, client, version.ID)
template := coderdtest.CreateTemplate(t, client, user.OrganizationID, version.ID)
workspace := coderdtest.CreateWorkspace(t, client, user.OrganizationID, template.ID)
build := coderdtest.AwaitWorkspaceBuildJob(t, client, workspace.LatestBuild.ID)
require.Equal(t, codersdk.WorkspaceStatusRunning, build.Status)
err = pd.Shutdown(ctx)
require.NoError(t, err)
err = terraformServer.Close()
require.NoError(t, err)
select {
case <-ctx.Done():
t.Error("timeout waiting for server to shut down")
case err := <-errCh:
require.NoError(t, err)
}
})
t.Run("BadPSK", func(t *testing.T) {
t.Parallel()
client, user := coderdenttest.New(t, &coderdenttest.Options{