feat: Rest endoint /health for rln (#2011)

* Rest endoint /health added

* Remove dev-debug echo

* Changed not ready message

* Update waku/node/rest/health/handlers.nim

Co-authored-by: Ivan Folgueira Bande <128452529+Ivansete-status@users.noreply.github.com>

* Various fixes on PR foundings, added comments

---------

Co-authored-by: Ivan Folgueira Bande <128452529+Ivansete-status@users.noreply.github.com>

apps/wakunode2/app.nim

@@ -47,6 +47,7 @@ import
   ../../waku/node/rest/relay/topic_cache,
   ../../waku/node/rest/filter/handlers as rest_filter_api,
   ../../waku/node/rest/store/handlers as rest_store_api,
+  ../../waku/node/rest/health/handlers as rest_health_api,
   ../../waku/node/jsonrpc/admin/handlers as rpc_admin_api,
   ../../waku/node/jsonrpc/debug/handlers as rpc_debug_api,
   ../../waku/node/jsonrpc/filter/handlers as rpc_filter_api,
@@ -569,6 +570,9 @@ proc startRestServer(app: App, address: ValidIpAddress, port: Port, conf: WakuNo
   ## Debug REST API
   installDebugApiHandlers(server.router, app.node)
 
+  ## Health REST API
+  installHealthApiHandler(server.router, app.node)
+
   ## Relay REST API
   if conf.relay:
     let relayCache = TopicCache.init(capacity=conf.restRelayCacheCapacity)

tests/all_tests_waku.nim

@@ -102,4 +102,5 @@ when defined(rln):
     ./waku_rln_relay/test_waku_rln_relay,
     ./waku_rln_relay/test_wakunode_rln_relay,
     ./waku_rln_relay/test_rln_group_manager_onchain,
-    ./waku_rln_relay/test_rln_group_manager_static
+    ./waku_rln_relay/test_rln_group_manager_static,
+    ./wakunode_rest/test_rest_health

tests/wakunode_rest/test_rest_health.nim

@@ -0,0 +1,77 @@
+{.used.}
+
+import
+  std/tempfiles,
+  stew/shims/net,
+  testutils/unittests,
+  presto,
+  presto/client as presto_client,
+  libp2p/peerinfo,
+  libp2p/multiaddress,
+  libp2p/crypto/crypto
+import
+  ../../waku/waku_node,
+  ../../waku/node/waku_node as waku_node2,  # TODO: Remove after moving `git_version` to the app code.
+  ../../waku/node/rest/server,
+  ../../waku/node/rest/client,
+  ../../waku/node/rest/responses,
+  ../../waku/node/rest/health/handlers as health_api,
+  ../../waku/node/rest/health/client as health_api_client,
+  ../../waku/waku_rln_relay,
+  ../testlib/common,
+  ../testlib/wakucore,
+  ../testlib/wakunode
+
+
+proc testWakuNode(): WakuNode =
+  let
+    privkey = crypto.PrivateKey.random(Secp256k1, rng[]).tryGet()
+    bindIp = ValidIpAddress.init("0.0.0.0")
+    extIp = ValidIpAddress.init("127.0.0.1")
+    port = Port(0)
+
+  newTestWakuNode(privkey, bindIp, port, some(extIp), some(port))
+
+
+suite "Waku v2 REST API - health":
+  asyncTest "Get node health info - GET /health":
+    # Given
+    let node = testWakuNode()
+    await node.start()
+    await node.mountRelay()
+
+    let restPort = Port(58001)
+    let restAddress = ValidIpAddress.init("0.0.0.0")
+    let restServer = RestServerRef.init(restAddress, restPort).tryGet()
+
+    installHealthApiHandler(restServer.router, node)
+    restServer.start()
+    let client = newRestHttpClient(initTAddress(restAddress, restPort))
+
+    # When
+    var response = await client.healthCheck()
+
+    # Then
+    check:
+      response.status == 503
+      $response.contentType == $MIMETYPE_TEXT
+      response.data == "Node is not ready"
+
+    # now kick in rln (currently the only check for health)
+    await node.mountRlnRelay(WakuRlnConfig(rlnRelayDynamic: false,
+      rlnRelayCredIndex: some(1.uint),
+      rlnRelayTreePath: genTempPath("rln_tree", "wakunode"),
+    ))
+
+    # When
+    response = await client.healthCheck()
+
+    # Then
+    check:
+      response.status == 200
+      $response.contentType == $MIMETYPE_TEXT
+      response.data == "Node is healthy"
+
+    await restServer.stop()
+    await restServer.closeWait()
+    await node.stop()

waku/node/rest/health/client.nim

@@ -0,0 +1,30 @@
+when (NimMajor, NimMinor) < (1, 4):
+  {.push raises: [Defect].}
+else:
+  {.push raises: [].}
+
+import
+  chronicles,
+  json_serialization,
+  json_serialization/std/options,
+  presto/[route, client]
+import
+  ../serdes,
+  ../responses
+
+logScope:
+  topics = "waku node rest health_api"
+
+proc decodeBytes*(t: typedesc[string], value: openArray[byte],
+                  contentType: Opt[ContentTypeData]): RestResult[string] =
+  if MediaType.init($contentType) != MIMETYPE_TEXT:
+    error "Unsupported contentType value", contentType = contentType
+    return err("Unsupported contentType")
+
+  var res: string
+  if len(value) > 0:
+    res = newString(len(value))
+    copyMem(addr res[0], unsafeAddr value[0], len(value))
+  return ok(res)
+
+proc healthCheck*(): RestResponse[string] {.rest, endpoint: "/health", meth: HttpMethod.MethodGet.}

waku/node/rest/health/handlers.nim

@@ -0,0 +1,42 @@
+when (NimMajor, NimMinor) < (1, 4):
+  {.push raises: [Defect].}
+else:
+  {.push raises: [].}
+
+import
+  chronicles,
+  json_serialization,
+  presto/route
+import
+  ../../waku_node,
+  ../responses,
+  ../serdes
+
+logScope:
+  topics = "waku node rest health_api"
+
+const ROUTE_HEALTH* = "/health"
+
+const FutIsReadyTimout = 5.seconds
+
+proc installHealthApiHandler*(router: var RestRouter, node: WakuNode) =
+  ## /health endpoint provides information about node readiness to caller.
+  ## Currently it is restricted to checking RLN (if mounted) proper setup
+  ## TODO: Leter to extend it to a broader information about each subsystem state
+  ## report. Rest response to change to JSON structure that can hold exact detailed 
+  ## information.
+  
+  router.api(MethodGet, ROUTE_HEALTH) do () -> RestApiResponse:
+
+    let isReadyStateFut = node.isReady()
+    if not await isReadyStateFut.withTimeout(FutIsReadyTimout):
+       return RestApiResponse.internalServerError("Health check timed out")
+
+    var msg = "Node is healthy"
+    var status = Http200
+
+    if not isReadyStateFut.read():
+      msg = "Node is not ready"
+      status = Http503
+
+    return RestApiResponse.textResponse(msg, status)

waku/node/rest/health/openapi.yaml

@@ -0,0 +1,41 @@
+openapi: 3.0.3
+info:
+  title: Waku V2 node REST API
+  version: 1.0.0
+  contact:
+    name: VAC Team
+    url: https://forum.vac.dev/
+
+tags:
+  - name: health
+    description: Healt check REST API for WakuV2 node
+
+paths:
+  /health:
+    get:
+      summary: Get node health status
+      description: Retrieve readiness of a Waku v2 node.
+      operationId: healthcheck
+      tags:
+        - health
+      responses:
+        '200':
+          description: Waku v2 node is up and running.
+          content:
+            text/plain:
+                schema:
+                  type: string
+                  example: Node is healty
+        '500':
+          description: Internal server error
+          content:
+            text/plain:
+                schema:
+                  type: string
+        '503':
+          description: Node not initialized or having issues
+          content:
+            text/plain:
+                schema:
+                  type: string
+                  example: Node is not initialized