> ## Documentation Index > Fetch the complete documentation index at: https://www.ravion.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # Start a rollback deployment > Start a rollback: create a new deployment that redeploys the exact task definition revision a previously-successful deploy registered — no new revision is created. The target deployment ID is in the URL path. Description is auto-generated. ## OpenAPI ````yaml https://api.ravion.com/openapi.yaml post /deployments/{id}/rollback openapi: 3.0.0 info: title: Ravion version: 0.0.0 servers: - url: https://api.ravion.com security: - BearerAuth: [] tags: - name: Projects - name: Environments - name: Pipelines - name: PipelineRuns - name: TerraformResources - name: TerraformExecutionSummaries - name: PipelineStepExecutions - name: AwsCloudWatch - name: PipelineVersions - name: Organizations - name: Stacks - name: StackWorkspaces - name: Auth - name: OAuth - name: User - name: Health - name: Memberships - name: ServiceAccounts - name: AwsDefaultNetworks - name: AwsAccounts - name: ApiKeys - name: ExecutionEnvironments - name: ModuleDefinitions - name: ModuleVersions - name: ModuleInstances - name: DefaultValueDefinitions - name: DefaultValues - name: CodeSources - name: Github - name: Gitlab - name: Git - name: Values - name: Deployments - name: DeploymentResources - name: InfrastructureEvents - name: WebSocket - name: Describe paths: /deployments/{id}/rollback: post: tags: - Deployments summary: Start a rollback deployment description: |- Start a rollback: create a new deployment that redeploys the exact task definition revision a previously-successful deploy registered — no new revision is created. The target deployment ID is in the URL path. Description is auto-generated. operationId: RollbackDeployment parameters: - in: path name: id required: true schema: type: string responses: '201': content: application/json: schema: additionalProperties: false properties: data: $ref: '#/components/schemas/Deployment.Model' required: - data type: object description: >- The request has succeeded and a new resource has been created as a result. '400': content: application/json: schema: $ref: '#/components/schemas/Errors.UserFacingErrorData' description: The request is invalid or malformed. '401': content: application/json: schema: $ref: '#/components/schemas/Errors.UserFacingErrorData' description: You are not authenticated '403': content: application/json: schema: $ref: '#/components/schemas/Errors.UserFacingErrorData' description: You do not have permission to access this resource. '404': content: application/json: schema: $ref: '#/components/schemas/Errors.UserFacingErrorData' description: The server cannot find the requested resource. '422': content: application/json: schema: $ref: '#/components/schemas/Errors.UserFacingErrorData' description: Client error '500': content: application/json: schema: $ref: '#/components/schemas/Errors.UserFacingErrorData' description: Server error components: schemas: Deployment.Model: additionalProperties: false properties: completedAt: allOf: - $ref: '#/components/schemas/UtcDateTime' description: Timestamp when the deployment completed readOnly: true createdAt: allOf: - $ref: '#/components/schemas/UtcDateTime' readOnly: true data: allOf: - $ref: '#/components/schemas/DeploymentData' description: |- Typed deployment-type-specific metadata (clusterArn / serviceArn / TD snapshots for ECS; functionArn / aliasName / version for Lambda; etc.). Stored as a JSON column on the ModuleDeployment row; payload shape is defined in `@fc/schemas/deployments/tsp`. readOnly: true deploymentType: description: Deployment type (e.g., "aws:ecs") type: string description: description: User-provided description type: string environmentId: description: Environment ID type: string error: allOf: - $ref: '#/components/schemas/Errors.UserFacingErrorData' description: User-facing error data, formatted from raw ErrorData stored in DB readOnly: true externalDeploymentId: description: |- Cloud-assigned deployment identity. ECS uses values like `ecs-svc/9124318887254438708`; the deployment list uses this to match ECS service-revision ARNs back to Ravion deployment rows. readOnly: true type: string id: readOnly: true type: string moduleInstanceId: description: Module instance ID this deployment targets type: string organizationId: readOnly: true type: string pipelineRunId: description: Pipeline run ID executing this deployment type: string projectId: description: Project ID type: string redeployOfDeploymentId: description: |- Populated on redeploy deployments — references the source deployment whose task definition revision was redeployed. readOnly: true type: string resourceArn: description: |- Hot-path correlation key for event ingestion. ECS → serviceArn; Lambda → functionArn; Static → primary CloudFront distribution id. Indexed on the DB row so event ingesters can resolve an incoming event to a deploy without JSON_EXTRACT. readOnly: true type: string resourceName: description: Human-readable name derived from the resource ARN. readOnly: true type: string rollbackToDeploymentId: description: |- Populated on rollback deployments — references the deployment that was rolled back to. readOnly: true type: string startedAt: allOf: - $ref: '#/components/schemas/UtcDateTime' description: Timestamp when the deployment started executing readOnly: true status: description: Deployment status enum: - PENDING - IN_PROGRESS - SUCCESS - ERROR - CANCELLED - ROLLING_BACK - ROLLED_BACK readOnly: true type: string stepExecutionId: description: |- Step execution that produced this deployment. Set when triggerSource=PIPELINE — points at the exact `deploy`/`deploy-swap` step inside `pipelineRunId` so the UI can deep-link without guessing by stepType. type: string triggerSource: description: How the deployment was triggered type: string updatedAt: allOf: - $ref: '#/components/schemas/UtcDateTime' readOnly: true required: - id - createdAt - updatedAt - organizationId - moduleInstanceId - projectId - environmentId - description - status - deploymentType - triggerSource type: object Errors.UserFacingErrorData: additionalProperties: false description: |- User-facing error presentation data. This is what the API returns to the frontend after formatting ErrorData using CEL templates from the error registry. Used for both: - Error fields on domain models (e.g., PipelineRun.error) - API error response bodies (HTTP 4xx/5xx responses) properties: action: allOf: - $ref: '#/components/schemas/Errors.Action' description: Optional action to help resolve the error code: description: Full error code, e.g., "Ravion:Pipeline:NOT_FOUND" type: string description: description: Additional description with more details type: string details: description: Structured details rendered as user-facing sections. items: $ref: '#/components/schemas/Errors.UserFacingErrorDetailSection' type: array isInternal: description: >- Indicates whether this error is internal (only set when ShowInternal=true). This allows SUPERADMINs to identify internal errors while viewing full details. type: boolean message: description: Main error message (required) type: string metadata: additionalProperties: {} description: >- Error params/metadata. Stripped for internal errors unless superadmin. type: object requestId: description: Request ID for correlating errors with server logs. type: string required: - code - message type: object UtcDateTime: description: UTC RFC3339 timestamp with a required Z suffix format: date-time pattern: ^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}(\.\d{1,9})?Z$ type: string x-go-type: apitypes.UTCDateTime x-go-type-import: name: apitypes path: ravion/api-go/pkg/apitypes DeploymentData: description: |- Discriminated union over deployment-type metadata. Variants are added as deployment types land: `aws:ecs` → `Deployments.EcsDeploymentData` `aws:ec2` → `Deployments.AwsEc2DeploymentData` `aws:static` → `Deployments.AwsStaticDeploymentData` `aws:lambda` → `Deployments.AwsLambdaDeploymentData` Variants live in `@fc/schemas/deployments/tsp/aws-ecs.tsp` and `aws-static.tsp` / `aws-lambda.tsp`; this union wires them up with the shared `type` discriminator so consumers can narrow on `data.type` without parsing opaque JSON. discriminator: mapping: aws:ec2: $ref: '#/components/schemas/Deployments.AwsEc2DeploymentData' aws:ecs: $ref: '#/components/schemas/Deployments.EcsDeploymentData' aws:lambda: $ref: '#/components/schemas/Deployments.AwsLambdaDeploymentData' aws:static: $ref: '#/components/schemas/Deployments.AwsStaticDeploymentData' propertyName: type oneOf: - $ref: '#/components/schemas/Deployments.EcsDeploymentData' - $ref: '#/components/schemas/Deployments.AwsEc2DeploymentData' - $ref: '#/components/schemas/Deployments.AwsStaticDeploymentData' - $ref: '#/components/schemas/Deployments.AwsLambdaDeploymentData' type: object Errors.Action: additionalProperties: false description: Action to help user resolve the error properties: label: description: Button/link label text type: string url: description: URL to navigate to for resolution type: string required: - label - url type: object Errors.UserFacingErrorDetailSection: additionalProperties: false description: Structured user-facing error detail section. properties: items: description: List of detail values for this section. items: type: string type: array object: additionalProperties: {} description: Structured detail payload for object rendering. type: object render: description: Rendering hint for clients. Valid values are list or object. type: string title: description: Detail section title shown in the UI. type: string required: - title - render type: object Deployments.AwsEc2DeploymentData: additionalProperties: false description: |- Typed payload for `ModuleDeployment.deploymentData` when `deploymentType == AWS_EC2`. Mirrors `AwsLambdaDeploymentData` — all per-deploy state folded into this JSON column so adding a deployment type costs zero new SQL tables. Indexed columns the workflow also denormalizes onto `ModuleDeployment` (`resourceArn`, `resourceName`, `externalDeploymentId`): - `resourceArn` ← Auto Scaling Group ARN (when known) - `resourceName` ← `autoScalingGroupName` - `externalDeploymentId` ← SSM `commandId` (commit identity). properties: autoScalingGroupName: description: Name of the Auto Scaling Group the deploy targets. type: string awsAccountNumber: description: |- Twelve-digit AWS account number that owns the Auto Scaling Group. Combined with region and ASG name for unambiguous EventBridge correlation when the group ARN is not present. type: string commandCompletedAt: description: Set when `commandStatus` reaches `SUCCESS` or `FAILED`. format: date-time type: string commandFailureReason: description: |- AWS error message if `commandStatus == FAILED` before any invocation ran (SendCommand itself failed). Per-instance failures land on `instanceResults[].outputTail` instead. type: string commandId: description: |- SSM command ID returned by SendCommand. Acts as the deploy's commit identity for downstream tooling. type: string commandStatus: $ref: '#/components/schemas/Deployments.AwsEc2CommandStatus' deployedDefinition: allOf: - $ref: '#/components/schemas/Deployments.AwsEc2DeploymentDefinition' description: Snapshot of the resolved definition this deploy applied. instanceResults: description: |- The InService instance snapshot the command was sent to, with each instance's terminal invocation status. items: $ref: '#/components/schemas/Deployments.AwsEc2InstanceResult' type: array logGroupName: description: |- CloudWatch log group receiving app output. Streams are named `deployment//instance/`. Absent for deployments created from module versions that predate runtime logs. type: string maxConcurrency: description: MaxConcurrency the command was sent with (count or percentage). type: string maxErrors: description: MaxErrors the command was sent with (count or percentage). type: string previousDefinition: allOf: - $ref: '#/components/schemas/Deployments.AwsEc2DeploymentDefinition' description: |- Snapshot of the resolved definition from the last SUCCESS-status EC2 deploy on this module instance, if any. Null on the very first EC2 deploy. region: description: AWS region of the group and document. type: string source: allOf: - $ref: '#/components/schemas/Pipeline.GitSource' description: >- Optional deploy.source used by a manual deploy. The stored ref is the immutable commit resolved before SSM dispatch. ssmDocumentName: description: |- Name of the SSM command document the deploy ran: the convention-derived `-deploy` (both runtimes). type: string targetGroupArn: description: |- ARN of the service target group, when the service is load-balanced. Informational — the document owns drain and re-register. type: string type: enum: - aws:ec2 type: string required: - type - autoScalingGroupName - ssmDocumentName - region - awsAccountNumber - commandStatus type: object Deployments.EcsDeploymentData: additionalProperties: false description: |- Typed payload for ModuleDeployment.deploymentData when deploymentType == AWS_ECS. Merged in from the old EcsModuleDeployment table. clusterArn + serviceArn live here in the JSON; the serviceArn is also denormalized onto the indexed ModuleDeployment.resourceArn column so the event-ingestion hot path (aws.ecs events) can resolve a deploy by service ARN without JSON_EXTRACT. properties: acceptedPauseControl: allOf: - $ref: >- #/components/schemas/Deployments.EcsDeploymentAcceptedPauseControl description: |- Latest PAUSED lifecycle hook control accepted by the API. This is cleared implicitly by mismatch: when a newer PAUSED event becomes latest, its hook/stage/timestamp no longer matches this object and the UI shows controls again. clusterArn: type: string deployedTaskDefinition: allOf: - $ref: '#/components/schemas/Deployments.EcsTaskDefinitionSnapshot' description: |- Snapshot of the task definition this deploy registered. Populated once RegisterTaskDefinition succeeds. ecsModuleDeploymentId: description: |- Preserved during migration from the old EcsModuleDeployment table so legacy log / ticket / Temporal-payload references can still be traced. New code should use the parent ModuleDeployment.id as the deploy identifier. type: string elbListenerState: description: |- ALB listener topology, one entry per load balancer the service is attached to. Written by the deploy workflow's `CaptureEcsServiceListeners` activity at phase-start only (topology rarely changes mid-deploy). Replaces the old `ECS_SERVICE_LISTENERS` snapshot kind. items: $ref: '#/components/schemas/Deployments.EcsElbListenerState' type: array events: description: |- Append-only progress timeline. The deploy workflow writes one entry per phase transition via JSON_ARRAY_APPEND so the row update fires the existing ModuleDeployment WebSocket broadcast. The frontend reads the last entry as the current phase to drive the deploy progress bar items: $ref: '#/components/schemas/Deployments.EcsDeploymentEvent' type: array postDeployEnabled: description: |- True when the resolved deploy config carried a non-empty `post_deploy` hook. Symmetric to preDeployEnabled. type: boolean preDeployEnabled: description: |- True when the resolved deploy config carried a non-empty `pre_deploy` hook. Written once at deploy-start, before any hook task runs. Drives the UI's pre-deploy phase slot rendering. Reads default to `false` when absent (legacy deployments that predate the hook feature). type: boolean previousTaskDefinition: allOf: - $ref: '#/components/schemas/Deployments.EcsTaskDefinitionSnapshot' description: |- Snapshot of the task definition that was running before this deploy registered a new one. Null on the very first deploy of a module instance. revertExternalDeploymentId: description: |- AWS deployment id (`ecs-svc/`) of the rollback UpdateService issued by the rolling revert path. Distinct from ModuleDeployment.externalDeploymentId, which stays bound to the failed deploy's own AWS deployment. type: string revertRequestedAt: description: |- Stamped by the rolling-strategy failure path just before its snapshot-revert UpdateService call. The revert starts a brand-new AWS service deployment whose id isn't known until UpdateService returns, so this flag lets event ingestion attribute that deployment's first events to this deploy (as its rollback) even when they beat the identity write below — instead of minting a phantom "external" deployment row. format: date-time type: string revertServiceDeploymentArn: description: |- Service-deployment ARN of the rollback deployment started by the rolling revert path. Companion to revertExternalDeploymentId for events that only carry the ARN. type: string serviceArn: type: string serviceDeploymentArn: description: |- ARN of the AWS service deployment (arn:aws:ecs:…:service-deployment/cluster/service/id) driving this deploy. Populated for non-rolling strategies once the UpdateService call resolves it via ListServiceDeployments; used by StopServiceDeployment / ContinueServiceDeployment and for EventBridge correlation. type: string serviceDeploymentState: allOf: - $ref: '#/components/schemas/Deployments.EcsServiceDeploymentState' description: |- Latest ECS service-deployment state captured from DescribeServiceDeployments. This mirrors ECS's target/source service revisions and drives Live / Standby badges without inferring traffic roles from row order or old Ravion phase events. strategy: allOf: - $ref: '#/components/schemas/Deployments.EcsDeployStrategy' description: |- Deployment strategy this deploy ran with. Absent on legacy rows (which predate native-strategy support and were always rolling or dual-service blue/green). type: enum: - aws:ecs type: string required: - type - clusterArn - serviceArn type: object Deployments.AwsLambdaDeploymentData: additionalProperties: false description: |- Typed payload for `ModuleDeployment.deploymentData` when `deploymentType == AWS_LAMBDA`. Mirrors `AwsStaticDeploymentData` — all per-deploy state folded into this JSON column so adding a deployment type costs zero new SQL tables. Indexed columns the workflow also denormalizes onto `ModuleDeployment` (`resourceArn`, `resourceName`, `externalDeploymentId`): - `resourceArn` ← `functionArn` - `resourceName` ← `functionName` - `externalDeploymentId` ← `:` (commit identity for the deploy). properties: aliasArn: description: Alias ARN after the successful UpdateAlias call. type: string aliasFailureReason: description: |- AWS error message if `aliasUpdateStatus == FAILED`. Most common cause is RevisionId mismatch from a racing deploy. type: string aliasName: description: |- Alias the deploy flips after publishing the new version. Defaults to `live` when the module YAML omits it. type: string aliasNewRevisionId: description: RevisionId returned by the successful UpdateAlias call. type: string aliasPreviousRevisionId: description: |- RevisionId observed on the alias *before* the UpdateAlias call — the value passed as `RevisionId` precondition so concurrent deploys can't silently clobber each other. type: string aliasUpdateCompletedAt: description: Set when `aliasUpdateStatus` transitions to `SUCCESS` or `FAILED`. format: date-time type: string aliasUpdateStatus: $ref: '#/components/schemas/Deployments.AwsLambdaAliasUpdateStatus' deployedDefinition: allOf: - $ref: '#/components/schemas/Deployments.AwsLambdaDeploymentDefinition' description: |- Snapshot of the resolved definition this deploy applied. Populated once the workflow finishes resolving inputs (before UpdateFunctionCode). functionArn: description: Full ARN of the Lambda function the deploy targets. type: string functionName: description: Short function name parsed from the ARN. type: string previousDefinition: allOf: - $ref: '#/components/schemas/Deployments.AwsLambdaDeploymentDefinition' description: |- Snapshot of the resolved definition from the last SUCCESS-status Lambda deploy on this module instance, if any. Null on the very first Lambda deploy. Used by the deploy detail drawer to render the definition diff. previousPublishedVersion: description: |- Function version the alias was pointing at *before* this deploy — the rollback target. Captured by the `CaptureLambdaPreviousAlias` activity. Null when this is the first deploy that created the alias. type: string publishFailureReason: description: AWS error message if `publishStatus == FAILED`. type: string publishStatus: $ref: '#/components/schemas/Deployments.AwsLambdaPublishStatus' publishedVersion: description: |- Numeric version produced by PublishVersion (e.g. `"42"`). Acts as the deploy's commit identity for downstream tooling. type: string publishedVersionArn: description: |- Qualified ARN of the published version (`:`). Denormalized for the UI. type: string region: description: AWS region of the function. type: string s3ArtifactBucket: description: |- Optional S3 bucket the build step uploads the zip artifact into. Captured for audit even when the deploy itself targets an IMAGE-packaged function (a single module may switch between packaging modes across deploys). type: string type: enum: - aws:lambda type: string updateCompletedAt: description: Set when `updateStatus` transitions to `SUCCESS` or `FAILED`. format: date-time type: string updateFailureReason: description: AWS error message if `updateStatus == FAILED`. type: string updateRevisionId: description: |- RevisionId returned by UpdateFunctionCode — captures the function configuration revision the deploy actually wrote. Useful for cross-correlating with Lambda's own audit log. type: string updateStatus: $ref: '#/components/schemas/Deployments.AwsLambdaUpdateStatus' required: - type - functionArn - functionName - region - aliasName - updateStatus - publishStatus - aliasUpdateStatus type: object Deployments.AwsStaticDeploymentData: additionalProperties: false description: |- Typed payload for `ModuleDeployment.deploymentData` when `deploymentType == AWS_STATIC`. Replaces the originally-planned per-deploy `AwsStaticModuleDeployment` child table — the deployment-manager refactor folded all per-type data into this JSON column so adding a deployment type costs zero new SQL tables. Indexed columns the workflow also denormalizes onto `ModuleDeployment` (`resourceArn`, `resourceName`, `externalDeploymentId`): - `resourceArn` ← `kvsArn` (so KVS-event ingesters can resolve a deploy by KVS arn without JSON_EXTRACT) - `resourceName` ← `s3Bucket` (UI breadcrumb) - `externalDeploymentId` ← `kvsETag` (commit identity, audit trail for promotion). properties: deployedDefinition: allOf: - $ref: '#/components/schemas/Deployments.AwsStaticDeploymentDefinition' description: |- Snapshot of the resolved definition this deploy promoted. Populated once the workflow finishes resolving inputs (before KVS write). deployedKvsState: additionalProperties: type: string description: |- KVS contents this deploy promoted, captured directly by the PutKey activity from the same `(key, value)` pair it wrote to CloudFront. This is the source of truth for the "after" side of the deploy detail diff — the renderer doesn't synthesize it from the resolved definition's directory field. Populated when `kvsStatus` transitions to SUCCESS. type: object distributions: description: |- Distributions whose CloudFront function reads from `kvsArn`. Captured at deploy phase-start by the `CaptureStaticDistributions` activity, mirroring how `CaptureEcsServiceListeners` works for ECS. Drives the deploy detail UI's domains list. items: $ref: '#/components/schemas/Deployments.AwsCloudfrontDistributionInfo' type: array invalidations: description: |- One entry per distribution the deploy was asked to invalidate. Empty / omitted when the deploy's definition specified no `cloudfront_invalidation_paths`. Updated incrementally by the background invalidation activity — the deploy itself completes the moment KVS PutKey succeeds, regardless of invalidation state. items: $ref: '#/components/schemas/Deployments.AwsCloudfrontInvalidationRecord' type: array kvsArn: description: ARN of the CloudFront KeyValueStore holding the `active` key. type: string kvsCompletedAt: description: Set when `kvsStatus` transitions to `SUCCESS` or `FAILED`. format: date-time type: string kvsETag: description: |- ETag returned by the successful `PutKey` call (the `ETag` response header). Acts as the deploy's commit identity and is the value the *next* deploy's `If-Match` header must carry to satisfy KVS optimistic concurrency. Also denormalized to `ModuleDeployment.externalDeploymentId` for indexed lookups. type: string kvsFailureReason: description: |- AWS error message if `kvsStatus == FAILED`. Most common cause is ETag mismatch from a racing deploy. type: string kvsPreviousETag: description: |- ETag observed on the `active` key immediately before this deploy's PutKey attempt. Captured for audit so post-mortem tooling can reconstruct the optimistic-concurrency chain across concurrent deploys. type: string kvsStatus: $ref: '#/components/schemas/Deployments.AwsStaticKvsWriteStatus' previousDefinition: allOf: - $ref: '#/components/schemas/Deployments.AwsStaticDeploymentDefinition' description: |- Snapshot of the resolved definition from the last SUCCESS-status static deploy on this module instance, if any. Null on the very first static deploy. Used by the deploy detail drawer to render the definition diff in lieu of a task-definition diff. previousKvsState: additionalProperties: type: string description: |- KVS contents *before* this deploy's PutKey, copied at create time from the prior successful deploy's `deployedKvsState`. Today this is a single-key map (`{[active]: }`) — modeled as a Record so the diff viewer doesn't have to know which keys the workflow promotes, and so multi-key promotions land without a UI rework. Null on the very first static deploy for a module instance. type: object prunedDirectories: description: |- Directories the prune activity removed from the hosting bucket for this deploy. Bounded by `keepLatestSuccessfulDeployCount` on the resolved definition. Empty when the count threshold wasn't exceeded. items: type: string type: array s3Bucket: description: Hosting bucket name (resolved `s3_bucket`). type: string s3Region: description: |- Hosting bucket region — needed by the prune activity to talk to the right S3 endpoint. type: string type: enum: - aws:static type: string required: - type - kvsArn - s3Bucket - s3Region - kvsStatus type: object Deployments.AwsEc2CommandStatus: description: |- Status of the SendCommand dispatch + ListCommandInvocations poll loop as a whole. enum: - PENDING - IN_PROGRESS - SUCCESS - FAILED type: string Deployments.AwsEc2DeploymentDefinition: additionalProperties: false description: |- Snapshot of the resolved `Module.Ec2ModuleDeployment.definition` (from `packages/schemas/modules/tsp/deployment.tsp`) captured at deploy time. Stored twice on `AwsEc2DeploymentData` (the prior successful deploy's snapshot + this deploy's snapshot) so the deploy detail drawer can render the definition diff. Open `raw` field — the underlying module-side definition schema can grow without forcing a schema migration here. properties: commands: description: Shell commands the deploy ran on each instance, in order. items: type: string type: array imageUri: description: |- Full image URI the deploy resolved (repository + tag or `@sha256:` digest). Renders as the deploy's UI tag for container deploys. type: string raw: additionalProperties: {} description: |- Raw resolved-definition JSON, kept verbatim so future fields surface in the diff viewer without a schema migration here. type: object runtime: allOf: - $ref: '#/components/schemas/Deployments.AwsEc2Runtime' description: |- Which runtime this deploy used. Drives which release fields below carry values. required: - runtime - raw type: object Deployments.AwsEc2InstanceResult: additionalProperties: false description: One instance's command invocation result. properties: completedAt: description: Set when the invocation reached a terminal status. format: date-time type: string instanceId: description: EC2 instance ID the document ran on. type: string outputTail: description: |- Tail of the invocation's stdout/stderr, captured for failed invocations so the deploy error surfaces the on-instance cause without an extra AWS round trip. type: string status: allOf: - $ref: '#/components/schemas/Deployments.AwsEc2InstanceInvocationStatus' description: Invocation status. required: - instanceId - status type: object Pipeline.GitSource: additionalProperties: false description: Git source configuration for fetching code properties: base_path: default: . description: |- Base path within the repository to use as the working root Example: base_path: packages/api Optional — defaults to "." type: string x-templateable: true branch: description: |- Branch to check out Example: branch: main Required minLength: 1 type: string x-templateable: true ref: description: >- Git ref to check out — branch, commit SHA, tag, or other ref. Overrides branch HEAD if set. Example: ref: v1.2.3 Optional type: string x-templateable: true repo: allOf: - $ref: '#/components/schemas/Pipeline.GitRepoUrl' description: |- Full repository URL including host Example: repo: https://gitlab.com/my-org/my-group/my-repo Required x-templateable: true type: description: Source type identifier enum: - git type: string x-templateable: true required: - type - repo - branch type: object Deployments.EcsDeploymentAcceptedPauseControl: additionalProperties: false properties: acceptedAt: description: When the API accepted and persisted this operator decision. format: date-time type: string action: allOf: - $ref: '#/components/schemas/Deployments.EcsDeploymentPauseControlAction' description: Accepted operator decision. eventStartedAt: description: |- `startedAt` from the PAUSED event that was resolved, when known. Lets consumers distinguish repeated pauses that use the same stage. format: date-time type: string hookId: description: |- Hook id the decision was sent for. Mirrors the PAUSED event's `hookId`; empty string when the event/API request omitted one. type: string labelKind: allOf: - $ref: >- #/components/schemas/Deployments.EcsDeploymentAcceptedPauseControlLabelKind description: Backend-derived label intent for the persisted operator decision. lifecycleStage: description: |- Lifecycle stage from the PAUSED event that was resolved, when known. Used by the UI to avoid hiding controls for a later pause. type: string required: - action - hookId - acceptedAt type: object Deployments.EcsTaskDefinitionSnapshot: additionalProperties: false description: |- Snapshot of an ECS task definition captured at deploy time. Carries AWS's assigned ARN + revision plus the raw task-definition payload AWS returned. Stored twice on EcsDeploymentData (before / after the deploy's RegisterTaskDefinition call) so the drawer can diff them. `definition` is an open-shape map of the `DescribeTaskDefinition` JSON AWS returned, keys normalized to the camelCase shape the AWS public API uses. We deliberately do NOT coerce into `Module.EcsTaskDefinition` (the user YAML schema) here — that schema is a subset of the AWS surface, so coercion would silently drop fields AWS includes but users don't get to set directly (registeredAt, status, requiresAttributes, revision, etc.). properties: arn: description: |- Full ARN returned by RegisterTaskDefinition, e.g. arn:aws:ecs:us-east-1:123:task-definition/my-svc:42 type: string definition: additionalProperties: {} description: |- Raw AWS `DescribeTaskDefinition` response, keys normalized to camelCase. Every field AWS returns (containerDefinitions, networkMode, requiresCompatibilities, taskRoleArn, registeredAt, …) is present verbatim so the drawer's diff viewer can render the complete before / after. type: object revision: description: |- Revision number AWS assigned. Duplicated from the ARN tail so consumers that parse family:revision directly don't need to string-split the ARN. format: int32 type: integer required: - arn - revision - definition type: object Deployments.EcsElbListenerState: additionalProperties: false description: |- Listener topology for one load balancer. Stored on `EcsDeploymentData.elbListenerState` as one entry per LB the service is attached to. Captured once at phase-start — listeners change when the user edits the module, not on every deploy tick. properties: capturedAt: description: When this topology was captured. format: date-time type: string loadBalancerArn: type: string loadBalancerDnsName: type: string rules: description: |- One entry per (listener, rule) pair on the load balancer. The drawer filters by targetGroupArn to surface only the rules that route to the task's TG. items: $ref: '#/components/schemas/Deployments.EcsListenerRule' type: array required: - loadBalancerArn - loadBalancerDnsName - rules - capturedAt type: object Deployments.EcsDeploymentEvent: additionalProperties: false description: |- One progress entry in `EcsDeploymentData.events`. Appended atomically via JSON_ARRAY_APPEND from the deploy workflow at every phase transition; the row update fires the existing ModuleDeployment WebSocket broadcast so the dashboard updates without polling. The array is append-only; consumers read the last entry as the current phase, and the full array as the deploy timeline. properties: currentStep: description: 1-indexed current step. Used by the progress bar denominator. format: int32 type: integer desiredCount: description: |- ECS service desiredCount once the snapshot phase has run. Used by the frontend to render "running tasks: N/M" during the WAIT_HEALTHY phase. Omitted on events fired before snapshot. format: int32 type: integer hookId: description: |- Pending PAUSE hook id (PAUSED events only). The continue API echoes this back via `ContinueServiceDeployment`. type: string lifecycleStage: description: |- ECS lifecycle stage that produced this event, when known (TRAFFIC_SHIFT / BAKE / PAUSED events). AWS enum spelling, e.g. POST_TEST_TRAFFIC_SHIFT, PRE_PRODUCTION_TRAFFIC_SHIFT. type: string message: description: |- Free-form human message captured alongside terminal events (root-cause string on FAILED). Omitted on intermediate phases. type: string startedAt: description: When the workflow recorded this event. format: date-time type: string stepCode: allOf: - $ref: '#/components/schemas/Deployments.EcsDeployPhaseCode' description: Phase identifier — see EcsDeployPhaseCode. totalSteps: description: Total user-visible steps in the deploy. Constant per workflow. format: int32 type: integer trafficPhase: allOf: - $ref: '#/components/schemas/Deployments.EcsDeploymentTrafficPhase' description: |- Backend-classified production/standby meaning for this event, derived from `stepCode` and `lifecycleStage`. Omitted when the event has no traffic-role implication. required: - stepCode - currentStep - totalSteps - startedAt type: object Deployments.EcsServiceDeploymentState: additionalProperties: false description: |- Current ECS service-deployment snapshot captured from DescribeServiceDeployments. This is the source of truth for deployment list Live / Standby badges: target/source revisions and lifecycle stage are what ECS believes at observation time. properties: acceptsNativeStandbyRevert: description: |- True while ECS still accepts a native StopServiceDeployment call for reverting production traffic to the source revision. type: boolean lifecycleStage: description: |- AWS lifecycle stage, e.g. POST_TEST_TRAFFIC_SHIFT, PRODUCTION_TRAFFIC_SHIFT, BAKE_TIME, CLEAN_UP. type: string observedAt: format: date-time type: string serviceDeploymentArn: type: string sourceServiceRevisions: items: $ref: '#/components/schemas/Deployments.EcsServiceRevisionTrafficState' type: array status: description: |- AWS service deployment status, e.g. PENDING, IN_PROGRESS, SUCCESSFUL, ROLLBACK_IN_PROGRESS, STOPPED. type: string statusReason: type: string strategy: description: |- AWS deployment strategy spelling from deploymentConfiguration, e.g. ROLLING, BLUE_GREEN, LINEAR, CANARY. type: string targetServiceRevision: $ref: '#/components/schemas/Deployments.EcsServiceRevisionTrafficState' trafficPhase: allOf: - $ref: '#/components/schemas/Deployments.EcsDeploymentTrafficPhase' description: |- Backend-classified traffic phase for the lifecycle stage. Omitted when ECS reports an unknown/empty stage and traffic weights are not enough by themselves to classify the snapshot. required: - serviceDeploymentArn - status - acceptsNativeStandbyRevert - observedAt type: object Deployments.EcsDeployStrategy: description: |- Deployment strategy executed by the native ECS deployment controller. Mirrors `Module.EcsDeploymentStrategyConfig.type` — the user-facing module YAML values, not the AWS enum spelling. enum: - rolling - blue_green - linear - canary type: string Deployments.AwsLambdaAliasUpdateStatus: description: |- Status of `UpdateAlias` — the commit point that flips traffic onto the newly published version. enum: - PENDING - SUCCESS - FAILED type: string Deployments.AwsLambdaDeploymentDefinition: additionalProperties: false description: |- Snapshot of the resolved `Module.LambdaModuleDeployment.definition` (from `packages/schemas/modules/tsp/deployment.tsp`) captured at deploy time. Stored twice on `AwsLambdaDeploymentData` (the prior successful deploy's snapshot + this deploy's snapshot) so the deploy detail drawer can render the definition diff. Open `raw` field — the underlying module-side definition schema can grow without forcing a schema migration here. properties: architectures: description: CPU architecture(s) — single-element array on AWS today. items: type: string type: array deadLetterTargetArn: description: Dead-letter target ARN (SQS queue or SNS topic). type: string description: description: Function description. type: string environment: additionalProperties: type: string description: |- Environment variables applied to the function. Replaces — Lambda doesn't merge. type: object ephemeralStorageSize: description: Ephemeral `/tmp` storage size in MiB. format: int32 type: integer fileSystemConfigs: description: EFS file system mounts. items: $ref: '#/components/schemas/Deployments.AwsLambdaFileSystemConfigSnapshot' type: array handler: description: Handler entrypoint (e.g. `index.handler`) — Zip only. type: string imageConfig: allOf: - $ref: '#/components/schemas/Deployments.AwsLambdaImageConfigSnapshot' description: Image-config overrides — Image-packaged functions only. imageUri: description: |- Full ECR image URI (with `@sha256:` digest preferred over a mutable tag — the workflow records the resolved digest where possible so the deploy is reproducible). type: string kmsKeyArn: description: KMS key ARN for environment variable encryption. type: string layers: description: Lambda layer ARNs (versioned). items: type: string type: array loggingConfig: allOf: - $ref: '#/components/schemas/Deployments.AwsLambdaLoggingConfigSnapshot' description: CloudWatch logging configuration. memorySize: description: Memory size in MiB. format: int32 type: integer packageType: allOf: - $ref: '#/components/schemas/Deployments.AwsLambdaPackageType' description: |- Which packaging mode this deploy used. Drives which of the zip- vs image-prefixed fields below carry values. raw: additionalProperties: {} description: |- Raw resolved-definition + configuration JSON, kept verbatim so future fields surface in the diff viewer without a schema migration here. type: object role: description: IAM execution role ARN. type: string runtime: description: Lambda runtime (e.g. `nodejs20.x`, `python3.12`) — Zip only. type: string s3Bucket: description: |- S3 bucket holding the zip artifact (resolved `infrastructure.s3_artifact_bucket`, denormalized into the per-deploy snapshot so each deploy's exact source is reconstructable from this JSON alone). type: string s3Key: description: |- S3 key the deploy uploaded into. Renders as the deploy's UI tag for Zip deploys. type: string snapStartApplyOn: description: SnapStart mode (`None` | `PublishedVersions`). type: string timeout: description: |- Wall-clock timeout in seconds (Lambda function-level, distinct from the deploy-level `timeout` on the module deployment). format: int32 type: integer tracingMode: description: X-Ray tracing mode (`Active` | `PassThrough`). type: string vpcSecurityGroupIds: description: VPC security group IDs attached to the function's ENIs. items: type: string type: array vpcSubnetIds: description: VPC subnet IDs the function's ENIs are placed into. items: type: string type: array required: - packageType - raw type: object Deployments.AwsLambdaPublishStatus: description: |- Status of `PublishVersion` — mints a new immutable numbered version of the function pointing at the just-uploaded code + configuration. enum: - PENDING - SUCCESS - FAILED type: string Deployments.AwsLambdaUpdateStatus: description: |- Status of `UpdateFunctionCode` + the subsequent `WaitForLambdaUpdated` poll loop. UpdateFunctionCode itself returns synchronously, but `LastUpdateStatus` on the function config is what tells us the new package is actually live. enum: - PENDING - IN_PROGRESS - SUCCESS - FAILED type: string Deployments.AwsStaticDeploymentDefinition: additionalProperties: false description: |- Snapshot of the resolved `Module.AwsStaticDeploymentDefinition` (from `packages/schemas/modules/tsp/deployment.tsp`) captured at deploy time. Stored twice on `AwsStaticDeploymentData` (the prior successful deploy's snapshot + this deploy's snapshot) so the deploy detail drawer can render the definition diff the user requested. Open shape because the underlying module-side definition schema can grow — capturing the raw JSON keeps existing snapshots renderable when new optional fields land. properties: cloudfrontInvalidationPaths: description: |- Optional CloudFront invalidation paths the user requested for this deploy. Empty array means "no invalidation". items: type: string type: array directory: description: |- Resolved `s3_directory` value — the directory under the hosting bucket that holds this deploy's built artifacts. Pulled to a named field because it doubles as the deploy's UI tag. type: string keepLatestSuccessfulDeployCount: description: |- How many successful static deploys to retain in the hosting bucket before pruning older directories. Null means "keep all". format: int32 type: integer raw: additionalProperties: {} description: |- Raw resolved-definition JSON, kept verbatim so future fields on `AwsStaticDeploymentDefinition` surface in the diff viewer without a schema migration here. type: object required: - directory - raw type: object Deployments.AwsCloudfrontDistributionInfo: additionalProperties: false description: |- One CloudFront distribution attached to this hosting bucket. Captured at deploy phase-start so the deploy detail's "details table" can list every domain alias without round-tripping CloudFront on render. properties: defaultDomain: description: |- AWS-assigned default CloudFront domain (e.g. `d111111abcdef8.cloudfront.net`). Always present — useful as a fallback when the user hasn't configured an alias. type: string distributionArn: description: |- Full ARN, e.g. arn:aws:cloudfront::123456789012:distribution/E1ABCDEF. type: string distributionId: description: |- Short id parsed from the ARN tail (e.g. `E1ABCDEF`). Duplicated so the UI can render the id without splitting the ARN. type: string domainAliases: description: |- All domain aliases configured on the distribution (`Aliases.Items` from `GetDistributionConfig`). Drives the "domains" row in the deploy detail UI. items: type: string type: array required: - distributionArn - distributionId - domainAliases type: object Deployments.AwsCloudfrontInvalidationRecord: additionalProperties: false description: |- One CloudFront invalidation issued by this deploy. Populated by the background invalidation activity, one entry per distribution-arn the deploy was asked to invalidate. properties: completedAt: description: Set when `status` transitions out of `IN_PROGRESS`. format: date-time type: string distributionArn: description: Distribution this invalidation targets. type: string failureReason: description: AWS error message if `status == FAILED`. type: string invalidationId: description: |- AWS invalidation id from `CreateInvalidation` (e.g. `I2J0MWVNIHERZJ`). Null if the call itself failed before AWS assigned an id. type: string paths: description: |- Paths submitted to CreateInvalidation — usually `["/*"]` but callers can scope narrower. Captured for audit. items: type: string type: array startedAt: format: date-time type: string status: $ref: '#/components/schemas/Deployments.AwsCloudfrontInvalidationStatus' required: - distributionArn - paths - status - startedAt type: object Deployments.AwsStaticKvsWriteStatus: description: |- Status of writing the resolved directory value to the CloudFront KeyValueStore `active` key. PutKey is the deploy's commit point — once `SUCCESS`, traffic flips to the promoted directory immediately. enum: - PENDING - SUCCESS - FAILED type: string Deployments.AwsEc2Runtime: description: |- Which runtime the deploy targets — drives which release fields on `AwsEc2DeploymentDefinition` carry values. enum: - CONTAINER - MANUAL type: string Deployments.AwsEc2InstanceInvocationStatus: description: |- Terminal + in-flight states of one instance's command invocation. Mirrors the SSM invocation status value space. enum: - PENDING - IN_PROGRESS - SUCCESS - FAILED - TIMED_OUT - CANCELLED type: string Pipeline.GitRepoUrl: description: |- Full git repository URL including host. Examples: - https://github.com/my-org/my-repo - github.com/my-org/my-repo - https://github.com/my-org/my-repo.git - git minLength: 1 pattern: >- ^((https?://)?[a-zA-Z0-9][a-zA-Z0-9.-]*\.[a-zA-Z]{2,}/[a-zA-Z0-9._-]+(/[a-zA-Z0-9._-]+)+|git@[a-zA-Z0-9][a-zA-Z0-9.-]*\.[a-zA-Z]{2,}:[a-zA-Z0-9._-]+(/[a-zA-Z0-9._-]+)+)(\.git)?$ type: string Deployments.EcsDeploymentPauseControlAction: description: |- Operator action accepted for a paused ECS lifecycle hook. Stored outside the append-only `events` timeline so accepting a hook does not mutate historical PAUSED events, while still giving every dashboard session a persisted source of truth after refresh. enum: - CONTINUE - ROLLBACK type: string Deployments.EcsDeploymentAcceptedPauseControlLabelKind: description: |- UI label intent for an accepted pause control. The backend derives this from the operator action plus the AWS lifecycle stage that was resolved. enum: - continue - traffic_shift - revert_standby type: string Deployments.EcsListenerRule: additionalProperties: false description: |- One ALB listener rule. A single DescribeListeners/DescribeRules capture often produces many rows — the same listener can have multiple rules forwarding to different target groups (e.g. `/api/*` → service A, `/admin/*` → service B). properties: hostHeaders: description: host-header condition values. Empty for the default rule. items: type: string type: array isDefaultRule: type: boolean listenerArn: type: string pathPatterns: description: path-pattern condition values. Empty for the default rule. items: type: string type: array port: format: int32 type: integer protocol: type: string ruleArn: type: string rulePriority: description: |- Numeric priority of the rule. Null for the listener's default rule (which has no priority). format: int32 type: integer targetGroupArn: description: |- Target group this rule forwards to. Used by the drawer to pick the rules relevant to a given task (whose TG is resolved from the task's slot). type: string targetGroupName: description: |- Short target-group name for UI display (derived from the ARN tail). Optional because older rows predate the capture activity that populates it. type: string required: - listenerArn - port - protocol - ruleArn - isDefaultRule - targetGroupArn type: object Deployments.EcsDeployPhaseCode: description: |- Phase of an ECS deploy run, persisted on each progress event the workflow appends to `EcsDeploymentData.events`. The frontend branches on this code (not the human-readable label) to render step UI; label rendering lives entirely in the frontend. enum: - INIT - PROVISION - SNAPSHOT - PRE_DEPLOY - UPDATE - WAIT_HEALTHY - POST_DEPLOY - ACTIVATE - COMPLETE - FAILED - TRAFFIC_SHIFT - BAKE - PAUSED - ROLLBACK type: string Deployments.EcsDeploymentTrafficPhase: description: |- Backend-classified traffic role for an ECS deployment event or DescribeServiceDeployments snapshot. This keeps AWS lifecycle-stage bucket logic on the backend, where the raw ECS status model is interpreted, instead of duplicating those string sets in the UI. enum: - standby - production type: string Deployments.EcsServiceRevisionTrafficState: additionalProperties: false description: |- One ECS service revision entry from DescribeServiceDeployments. ECS reports target and source revisions in terms of service-revision ARNs rather than Ravion deployment IDs; consumers match the ARN tail to `ModuleDeployment.externalDeploymentId` (`ecs-svc/`) to identify the corresponding row. properties: pendingTaskCount: format: int32 type: integer requestedProductionTrafficWeight: description: |- Percentage of production traffic ECS reports for this revision. Optional because AWS omits it for deployment shapes that do not expose traffic weights. format: double type: number requestedTaskCount: format: int32 type: integer requestedTestTrafficWeight: description: |- Percentage of test traffic ECS reports for this revision. Optional for the same reason as requestedProductionTrafficWeight. format: double type: number runningTaskCount: format: int32 type: integer serviceRevisionArn: description: |- ARN such as arn:aws:ecs:us-east-2:123:service-revision/cluster/service/912... type: string required: - serviceRevisionArn - requestedTaskCount - runningTaskCount - pendingTaskCount type: object Deployments.AwsLambdaFileSystemConfigSnapshot: additionalProperties: false description: EFS mount snapshot. properties: arn: type: string localMountPath: type: string required: - arn - localMountPath type: object Deployments.AwsLambdaImageConfigSnapshot: additionalProperties: false description: Image-config snapshot — Image-packaged functions only. properties: command: items: type: string type: array entryPoint: items: type: string type: array workingDirectory: type: string type: object Deployments.AwsLambdaLoggingConfigSnapshot: additionalProperties: false description: Logging configuration snapshot. properties: applicationLogLevel: type: string logFormat: type: string logGroup: type: string systemLogLevel: type: string type: object Deployments.AwsLambdaPackageType: description: |- Which packaging mode the deploy targets — drives which code-source fields on `AwsLambdaDeploymentDefinition` carry values. enum: - ZIP - IMAGE type: string Deployments.AwsCloudfrontInvalidationStatus: description: Status of an individual CloudFront `CreateInvalidation` call. enum: - IN_PROGRESS - COMPLETED - FAILED type: string securitySchemes: BearerAuth: scheme: Bearer type: http ````