Files
tiny-forge/internal/deployer/bluegreen.go
T
alexei.dolgolyov be6ad15efc fix: comprehensive security, performance, and quality hardening
Security: apply AdminOnly middleware to mutating routes, require
ENCRYPTION_KEY and ADMIN_PASSWORD (no insecure defaults), restrict
CORS to same-origin, fix OIDC token delivery via cookie instead of
URL query param, add rate limiting on login, add MaxBytesReader,
validate volume paths against traversal, add security headers,
validate user roles, add Secure flag to OIDC cookie.

Performance: set SQLite MaxOpenConns(1) to prevent SQLITE_BUSY,
add FK indexes on 8 columns, track notifier goroutines with
WaitGroup for graceful shutdown, use GetRegistryByName instead of
GetAllRegistries in deployer, pass basePath param to avoid redundant
settings query, return empty slices from store to remove reflection.

Quality: refactor TriggerDeploy to delegate to runDeploy (~100 lines
removed), consolidate duplicated utilities (extractPort, boolToInt,
now, isTerminalStatus) into shared exports, migrate all log.Printf
to slog structured logging, use consistent webhook response envelope,
remove dead code (parseEnvVars, duplicate auth types).

UX: clean up NPM proxy on instance removal via API, add README with
quickstart guide, add .env.example, require ADMIN_PASSWORD in
docker-compose, document staging-net prerequisite.
2026-03-29 12:49:24 +03:00

176 lines
6.0 KiB
Go

package deployer
import (
"context"
"fmt"
"log/slog"
"github.com/alexei/docker-watcher/internal/docker"
"github.com/alexei/docker-watcher/internal/store"
"github.com/google/uuid"
)
// blueGreenDeploy performs a zero-downtime deployment:
// 1. Start new container (green)
// 2. Health check green
// 3. Swap NPM proxy to point to green
// 4. Stop old container (blue)
//
// If the new container fails health check, it is removed and the old one stays.
func (d *Deployer) blueGreenDeploy(
ctx context.Context,
project store.Project,
stage store.Stage,
settings store.Settings,
deployID string,
imageTag string,
) (string, int, string, error) {
// Find existing running instance for this stage (the "blue" instance).
existingInstances, err := d.store.GetInstancesByStageID(stage.ID)
if err != nil {
return "", 0, "", fmt.Errorf("get existing instances: %w", err)
}
var blueInstance *store.Instance
for _, inst := range existingInstances {
if inst.Status == "running" {
instCopy := inst
blueInstance = &instCopy
break
}
}
// Step 1: Pull image.
if err := d.store.UpdateDeployStatus(deployID, "pulling", ""); err != nil {
slog.Warn("update deploy status", "error", err)
}
d.publishDeployStatus(deployID, project.ID, stage.ID, imageTag, "pulling", "")
d.logDeploy(deployID, fmt.Sprintf("Blue-green: pulling image %s:%s", project.Image, imageTag), "info")
authConfig, err := d.buildRegistryAuth(project)
if err != nil {
return "", 0, "", fmt.Errorf("build registry auth: %w", err)
}
if err := d.docker.PullImage(ctx, project.Image, imageTag, authConfig); err != nil {
return "", 0, "", fmt.Errorf("pull image: %w", err)
}
d.logDeploy(deployID, "Image pulled successfully", "info")
// Step 2: Ensure network.
networkID, err := d.docker.EnsureNetwork(ctx, settings.Network)
if err != nil {
return "", 0, "", fmt.Errorf("ensure network: %w", err)
}
// Step 3: Create and start green container.
if err := d.store.UpdateDeployStatus(deployID, "starting", ""); err != nil {
slog.Warn("update deploy status", "error", err)
}
d.publishDeployStatus(deployID, project.ID, stage.ID, imageTag, "starting", "")
instanceID := uuid.New().String()
subdomain := d.buildSubdomain(project, stage, settings, imageTag)
containerName := docker.ContainerName(project.Name, stage.Name, imageTag)
portStr := fmt.Sprintf("%d/tcp", project.Port)
envVars := d.mergeEnvVars(project, stage.ID)
mounts := d.computeVolumeMounts(project.ID, stage.Name, imageTag, settings.BaseVolumePath)
containerCfg := docker.ContainerConfig{
Name: containerName,
Image: project.Image + ":" + imageTag,
Env: envVars,
ExposedPorts: []string{portStr},
NetworkName: settings.Network,
NetworkID: networkID,
Project: project.Name,
Stage: stage.Name,
InstanceID: instanceID,
Mounts: mounts,
}
d.logDeploy(deployID, fmt.Sprintf("Blue-green: creating green container %s", containerName), "info")
containerID, err := d.docker.CreateContainer(ctx, containerCfg)
if err != nil {
return "", 0, instanceID, fmt.Errorf("create container: %w", err)
}
// Create instance record.
inst, err := d.store.CreateInstanceWithID(store.Instance{
ID: instanceID,
StageID: stage.ID,
ProjectID: project.ID,
ContainerID: containerID,
ImageTag: imageTag,
Subdomain: subdomain,
Status: "stopped",
Port: project.Port,
})
if err != nil {
return containerID, 0, instanceID, fmt.Errorf("create instance record: %w", err)
}
instanceID = inst.ID
if err := d.store.SetDeployInstanceID(deployID, instanceID); err != nil {
slog.Warn("link deploy to instance", "error", err)
}
d.logDeploy(deployID, fmt.Sprintf("Blue-green: starting green container %s", containerName), "info")
if err := d.docker.StartContainer(ctx, containerID); err != nil {
return containerID, 0, instanceID, fmt.Errorf("start container: %w", err)
}
if err := d.store.UpdateInstanceStatus(instanceID, "running"); err != nil {
slog.Warn("update instance status", "error", err)
}
d.publishInstanceStatus(instanceID, project.ID, stage.ID, "running")
// Step 4: Health check the green container.
if project.Healthcheck != "" {
if err := d.store.UpdateDeployStatus(deployID, "health_checking", ""); err != nil {
slog.Warn("update deploy status", "error", err)
}
d.publishDeployStatus(deployID, project.ID, stage.ID, imageTag, "health_checking", "")
healthURL := fmt.Sprintf("http://%s:%d%s", containerName, project.Port, project.Healthcheck)
d.logDeploy(deployID, fmt.Sprintf("Blue-green: health checking green at %s", healthURL), "info")
if err := d.health.Check(ctx, healthURL); err != nil {
return containerID, 0, instanceID, fmt.Errorf("health check green: %w", err)
}
d.logDeploy(deployID, "Blue-green: green health check passed", "info")
}
// Step 5: Swap NPM proxy to green.
if err := d.store.UpdateDeployStatus(deployID, "configuring_proxy", ""); err != nil {
slog.Warn("update deploy status", "error", err)
}
d.publishDeployStatus(deployID, project.ID, stage.ID, imageTag, "configuring_proxy", "")
npmProxyID, err := d.configureProxy(ctx, deployID, settings, containerName, project.Port, subdomain)
if err != nil {
return containerID, 0, instanceID, fmt.Errorf("configure proxy: %w", err)
}
inst.NpmProxyID = npmProxyID
inst.Subdomain = subdomain
if err := d.store.UpdateInstance(inst); err != nil {
slog.Warn("update instance with proxy ID", "error", err)
}
d.logDeploy(deployID, "Blue-green: proxy swapped to green container", "info")
// Step 6: Stop the blue container.
if blueInstance != nil {
d.logDeploy(deployID, fmt.Sprintf("Blue-green: stopping blue instance %s (tag: %s)", blueInstance.ID, blueInstance.ImageTag), "info")
if err := d.removeInstance(ctx, *blueInstance, settings); err != nil {
// Non-fatal: log but continue. Green is already serving traffic.
d.logDeploy(deployID, fmt.Sprintf("Blue-green: warning: failed to remove blue instance: %v", err), "warn")
} else {
d.logDeploy(deployID, "Blue-green: blue instance removed", "info")
}
}
return containerID, npmProxyID, instanceID, nil
}