Fix minor issue in LCMHelmConn
[osm/LCM.git] / osm_lcm / lcm_helm_conn.py
index 555613e..34e4915 100644 (file)
@@ -20,21 +20,27 @@ import yaml
 import asyncio
 import socket
 import uuid
+import os
 
 from grpclib.client import Channel
 
 from osm_lcm.frontend_pb2 import PrimitiveRequest
 from osm_lcm.frontend_pb2 import SshKeyRequest, SshKeyReply
 from osm_lcm.frontend_grpc import FrontendExecutorStub
+from osm_lcm.lcm_utils import LcmBase
+
+from osm_lcm.data_utils.database.database import Database
+from osm_lcm.data_utils.filesystem.filesystem import Filesystem
 
 from n2vc.n2vc_conn import N2VCConnector
 from n2vc.k8s_helm_conn import K8sHelmConnector
+from n2vc.k8s_helm3_conn import K8sHelm3Connector
 from n2vc.exceptions import N2VCBadArgumentsException, N2VCException, N2VCExecutionException
 
 from osm_lcm.lcm_utils import deep_get
 
 
-def retryer(max_wait_time=60, delay_time=10):
+def retryer(max_wait_time_var="_initial_retry_time", delay_time_var="_retry_delay"):
     def wrapper(func):
         retry_exceptions = (
             ConnectionRefusedError
@@ -42,6 +48,17 @@ def retryer(max_wait_time=60, delay_time=10):
 
         @functools.wraps(func)
         async def wrapped(*args, **kwargs):
+            # default values for wait time and delay_time
+            delay_time = 10
+            max_wait_time = 300
+
+            # obtain arguments from variable names
+            self = args[0]
+            if self.__dict__.get(max_wait_time_var):
+                max_wait_time = self.__dict__.get(max_wait_time_var)
+            if self.__dict__.get(delay_time_var):
+                delay_time = self.__dict__.get(delay_time_var)
+
             wait_time = max_wait_time
             while wait_time > 0:
                 try:
@@ -56,21 +73,19 @@ def retryer(max_wait_time=60, delay_time=10):
     return wrapper
 
 
-class LCMHelmConn(N2VCConnector):
+class LCMHelmConn(N2VCConnector, LcmBase):
     _KUBECTL_OSM_NAMESPACE = "osm"
     _KUBECTL_OSM_CLUSTER_NAME = "_system-osm-k8s"
     _EE_SERVICE_PORT = 50050
 
-    # Time beetween retries
-    _EE_RETRY_DELAY = 10
     # Initial max retry time
-    _MAX_INITIAL_RETRY_TIME = 300
-    # Other retry time
+    _MAX_INITIAL_RETRY_TIME = 600
+    # Max retry time for normal operations
     _MAX_RETRY_TIME = 30
+    # Time beetween retries, retry time after a connection error is raised
+    _EE_RETRY_DELAY = 10
 
     def __init__(self,
-                 db: object,
-                 fs: object,
                  log: object = None,
                  loop: object = None,
                  url: str = None,
@@ -81,33 +96,58 @@ class LCMHelmConn(N2VCConnector):
         Initialize EE helm connector.
         """
 
+        self.db = Database().instance.db
+        self.fs = Filesystem().instance.fs
+
         # parent class constructor
         N2VCConnector.__init__(
             self,
-            db=db,
-            fs=fs,
             log=log,
             loop=loop,
             url=url,
             username=username,
             vca_config=vca_config,
             on_update_db=on_update_db,
+            db=self.db,
+            fs=self.fs
         )
 
         self.log.debug("Initialize helm N2VC connector")
+        self.log.debug("initial vca_config: {}".format(vca_config))
 
         # TODO - Obtain data from configuration
         self._ee_service_port = self._EE_SERVICE_PORT
 
         self._retry_delay = self._EE_RETRY_DELAY
-        self._max_retry_time = self._MAX_RETRY_TIME
-        self._initial_retry_time = self._MAX_INITIAL_RETRY_TIME
 
-        # initialize helm connector
-        self._k8sclusterhelm = K8sHelmConnector(
+        if self.vca_config and self.vca_config.get("eegrpcinittimeout"):
+            self._initial_retry_time = self.vca_config.get("eegrpcinittimeout")
+            self.log.debug("Initial retry time: {}".format(self._initial_retry_time))
+        else:
+            self._initial_retry_time = self._MAX_INITIAL_RETRY_TIME
+            self.log.debug("Applied default retry time: {}".format(self._initial_retry_time))
+
+        if self.vca_config and self.vca_config.get("eegrpctimeout"):
+            self._max_retry_time = self.vca_config.get("eegrpctimeout")
+            self.log.debug("Retry time: {}".format(self._max_retry_time))
+        else:
+            self._max_retry_time = self._MAX_RETRY_TIME
+            self.log.debug("Applied default retry time: {}".format(self._max_retry_time))
+
+        # initialize helm connector for helmv2 and helmv3
+        self._k8sclusterhelm2 = K8sHelmConnector(
             kubectl_command=self.vca_config.get("kubectlpath"),
             helm_command=self.vca_config.get("helmpath"),
             fs=self.fs,
+            db=self.db,
+            log=self.log,
+            on_update_db=None,
+        )
+
+        self._k8sclusterhelm3 = K8sHelm3Connector(
+            kubectl_command=self.vca_config.get("kubectlpath"),
+            helm_command=self.vca_config.get("helm3path"),
+            fs=self.fs,
             log=self.log,
             db=self.db,
             on_update_db=None,
@@ -125,7 +165,8 @@ class LCMHelmConn(N2VCConnector):
                                            total_timeout: float = None,
                                            config: dict = None,
                                            artifact_path: str = None,
-                                           vca_type: str = None) -> (str, dict):
+                                           vca_type: str = None,
+                                           *kargs, **kwargs) -> (str, dict):
         """
         Creates a new helm execution environment deploying the helm-chat indicated in the
         attifact_path
@@ -140,7 +181,7 @@ class LCMHelmConn(N2VCConnector):
         :param float total_timeout:
         :param dict config:  General variables to instantiate KDU
         :param str artifact_path:  path of package content
-        :param str vca_type:  Type of vca, not used as assumed of type helm
+        :param str vca_type:  Type of vca, must be type helm or helm-v3
         :returns str, dict: id of the new execution environment including namespace.helm_id
         and credentials object set to None as all credentials should be osm kubernetes .kubeconfig
         """
@@ -157,7 +198,9 @@ class LCMHelmConn(N2VCConnector):
                 message="artifact_path is mandatory", bad_args=["artifact_path"]
             )
 
-        # Validate artifact-path exists
+        # Validate artifact-path exists and sync path
+        from_path = os.path.split(artifact_path)[0]
+        self.fs.sync(from_path)
 
         # remove / in charm path
         while artifact_path.find("//") >= 0:
@@ -175,10 +218,13 @@ class LCMHelmConn(N2VCConnector):
         else:
             full_path = self.fs.path + "/" + helm_chart_path
 
+        while full_path.find("//") >= 0:
+            full_path = full_path.replace("//", "/")
+
         try:
             # Call helm conn install
             # Obtain system cluster id from database
-            system_cluster_uuid = self._get_system_cluster_id()
+            system_cluster_uuid = await self._get_system_cluster_id()
             # Add parameter osm if exist to global
             if config and config.get("osm"):
                 if not config.get("global"):
@@ -186,13 +232,30 @@ class LCMHelmConn(N2VCConnector):
                 config["global"]["osm"] = config.get("osm")
 
             self.log.debug("install helm chart: {}".format(full_path))
-            helm_id = await self._k8sclusterhelm.install(system_cluster_uuid, kdu_model=full_path,
-                                                         namespace=self._KUBECTL_OSM_NAMESPACE,
-                                                         params=config,
-                                                         db_dict=db_dict,
-                                                         timeout=progress_timeout)
-
-            ee_id = "{}.{}".format(self._KUBECTL_OSM_NAMESPACE, helm_id)
+            if vca_type == "helm":
+                helm_id = self._k8sclusterhelm2.generate_kdu_instance_name(
+                    db_dict=db_dict,
+                    kdu_model=full_path,
+                )
+                await self._k8sclusterhelm2.install(system_cluster_uuid, kdu_model=full_path,
+                                                    kdu_instance=helm_id,
+                                                    namespace=self._KUBECTL_OSM_NAMESPACE,
+                                                    params=config,
+                                                    db_dict=db_dict,
+                                                    timeout=progress_timeout)
+            else:
+                helm_id = self._k8sclusterhelm2.generate_kdu_instance_name(
+                    db_dict=db_dict,
+                    kdu_model=full_path,
+                )
+                await self._k8sclusterhelm3.install(system_cluster_uuid, kdu_model=full_path,
+                                                    kdu_instance=helm_id,
+                                                    namespace=self._KUBECTL_OSM_NAMESPACE,
+                                                    params=config,
+                                                    db_dict=db_dict,
+                                                    timeout=progress_timeout)
+
+            ee_id = "{}:{}.{}".format(vca_type, self._KUBECTL_OSM_NAMESPACE, helm_id)
             return ee_id, None
         except N2VCException:
             raise
@@ -201,7 +264,8 @@ class LCMHelmConn(N2VCConnector):
             raise N2VCException("Error deploying chart ee: {}".format(e))
 
     async def register_execution_environment(self, namespace: str, credentials: dict, db_dict: dict,
-                                             progress_timeout: float = None, total_timeout: float = None) -> str:
+                                             progress_timeout: float = None, total_timeout: float = None,
+                                             *kargs, **kwargs) -> str:
         # nothing to do
         pass
 
@@ -256,7 +320,7 @@ class LCMHelmConn(N2VCConnector):
 
         try:
             # Obtain ip_addr for the ee service, it is resolved by dns from the ee name by kubernetes
-            namespace, helm_id = self._get_ee_id_parts(ee_id)
+            version, namespace, helm_id = self._get_ee_id_parts(ee_id)
             ip_addr = socket.gethostbyname(helm_id)
 
             # Obtain ssh_key from the ee, this method will implement retries to allow the ee
@@ -306,7 +370,7 @@ class LCMHelmConn(N2VCConnector):
             params_dict = dict()
 
         try:
-            namespace, helm_id = self._get_ee_id_parts(ee_id)
+            version, namespace, helm_id = self._get_ee_id_parts(ee_id)
             ip_addr = socket.gethostbyname(helm_id)
         except Exception as e:
             self.log.error("Error getting ee ip ee: {}".format(e))
@@ -370,7 +434,13 @@ class LCMHelmConn(N2VCConnector):
         # nothing to be done
         pass
 
-    async def delete_execution_environment(self, ee_id: str, db_dict: dict = None, total_timeout: float = None):
+    async def delete_execution_environment(
+        self,
+        ee_id: str,
+        db_dict: dict = None,
+        total_timeout: float = None,
+        **kwargs,
+    ):
         """
         Delete an execution environment
         :param str ee_id: id of the execution environment to delete, included namespace.helm_id
@@ -393,13 +463,17 @@ class LCMHelmConn(N2VCConnector):
         try:
 
             # Obtain cluster_uuid
-            system_cluster_uuid = self._get_system_cluster_id()
+            system_cluster_uuid = await self._get_system_cluster_id()
 
             # Get helm_id
-            namespace, helm_id = self._get_ee_id_parts(ee_id)
+            version, namespace, helm_id = self._get_ee_id_parts(ee_id)
 
-            # Uninstall chart
-            await self._k8sclusterhelm.uninstall(system_cluster_uuid, helm_id)
+            # Uninstall chart, for backward compatibility we must assume that if there is no
+            # version it is helm-v2
+            if version == "helm-v3":
+                await self._k8sclusterhelm3.uninstall(system_cluster_uuid, helm_id)
+            else:
+                await self._k8sclusterhelm2.uninstall(system_cluster_uuid, helm_id)
             self.log.info("ee_id: {} deleted".format(ee_id))
         except N2VCException:
             raise
@@ -420,10 +494,11 @@ class LCMHelmConn(N2VCConnector):
         progress_timeout: float = None,
         total_timeout: float = None,
         config: dict = None,
+        *kargs, **kwargs
     ) -> str:
         pass
 
-    @retryer(max_wait_time=_MAX_INITIAL_RETRY_TIME, delay_time=_EE_RETRY_DELAY)
+    @retryer(max_wait_time_var="_initial_retry_time", delay_time_var="_retry_delay")
     async def _get_ssh_key(self, ip_addr):
         channel = Channel(ip_addr, self._ee_service_port)
         try:
@@ -434,13 +509,13 @@ class LCMHelmConn(N2VCConnector):
         finally:
             channel.close()
 
-    @retryer(max_wait_time=_MAX_INITIAL_RETRY_TIME, delay_time=_EE_RETRY_DELAY)
+    @retryer(max_wait_time_var="_initial_retry_time", delay_time_var="_retry_delay")
     async def _execute_config_primitive(self, ip_addr, params, db_dict=None):
         return await self._execute_primitive_internal(ip_addr, "config", params, db_dict=db_dict)
 
-    @retryer(max_wait_time=_MAX_RETRY_TIME, delay_time=_EE_RETRY_DELAY)
+    @retryer(max_wait_time_var="_max_retry_time", delay_time_var="_retry_delay")
     async def _execute_primitive(self, ip_addr, primitive_name, params, db_dict=None):
-        return await  self._execute_primitive_internal(ip_addr, primitive_name, params, db_dict=db_dict)
+        return await self._execute_primitive_internal(ip_addr, primitive_name, params, db_dict=db_dict)
 
     async def _execute_primitive_internal(self, ip_addr, primitive_name, params, db_dict=None):
 
@@ -486,17 +561,35 @@ class LCMHelmConn(N2VCConnector):
         except Exception as e:
             self.log.error("Error writing detailedStatus to database: {}".format(e))
 
-    def _get_system_cluster_id(self):
+    async def _get_system_cluster_id(self):
         if not self._system_cluster_id:
             db_k8cluster = self.db.get_one("k8sclusters", {"name": self._KUBECTL_OSM_CLUSTER_NAME})
-            k8s_hc_id = deep_get(db_k8cluster, ("_admin", "helm-chart", "id"))
+            k8s_hc_id = deep_get(db_k8cluster, ("_admin", "helm-chart-v3", "id"))
             if not k8s_hc_id:
-                self.log.error("osm system cluster has not been properly initialized for helm connector, "
-                               "helm-chart id is not defined")
-                raise N2VCException("osm system cluster has not been properly initialized for helm connector")
+                try:
+                    # backward compatibility for existing clusters that have not been initialized for helm v3
+                    cluster_id = db_k8cluster.get("_id")
+                    k8s_credentials = yaml.safe_dump(db_k8cluster.get("credentials"))
+                    k8s_hc_id, uninstall_sw = await self._k8sclusterhelm3.init_env(k8s_credentials,
+                                                                                   reuse_cluster_uuid=cluster_id)
+                    db_k8scluster_update = {"_admin.helm-chart-v3.error_msg": None,
+                                            "_admin.helm-chart-v3.id": k8s_hc_id,
+                                            "_admin.helm-chart-v3}.created": uninstall_sw,
+                                            "_admin.helm-chart-v3.operationalState": "ENABLED"}
+                    self.update_db_2("k8sclusters", cluster_id, db_k8scluster_update)
+                except Exception as e:
+                    self.log.error("error initializing helm-v3 cluster: {}".format(str(e)))
+                    raise N2VCException("K8s system cluster '{}' has not been initialized for helm-chart-v3".format(
+                        cluster_id))
             self._system_cluster_id = k8s_hc_id
         return self._system_cluster_id
 
     def _get_ee_id_parts(self, ee_id):
-        namespace, _, helm_id = ee_id.partition('.')
-        return namespace, helm_id
+        """
+        Parses ee_id stored at database that can be either 'version:namespace.helm_id' or only
+        namespace.helm_id for backward compatibility
+        If exists helm version can be helm-v3 or helm (helm-v2 old version)
+        """
+        version, _, part_id = ee_id.rpartition(':')
+        namespace, _, helm_id = part_id.rpartition('.')
+        return version, namespace, helm_id