X-Git-Url: https://osm.etsi.org/gitweb/?a=blobdiff_plain;f=osm_lcm%2Flcm.py;h=d967c03854436ea3b39770330a99d529ed74867d;hb=da96482b5a0f4e2ad433f9ddd8a951eb59b322ba;hp=892ca4120a42961a664e9682784239d4093b2b7d;hpb=9ab9594cd42c85c422ef0c8e2d87a895150ab55a;p=osm%2FLCM.git diff --git a/osm_lcm/lcm.py b/osm_lcm/lcm.py index 892ca41..d967c03 100644 --- a/osm_lcm/lcm.py +++ b/osm_lcm/lcm.py @@ -1,6 +1,22 @@ #!/usr/bin/python3 # -*- coding: utf-8 -*- +## +# Copyright 2018 Telefonica S.A. +# +# Licensed under the Apache License, Version 2.0 (the "License"); you may +# not use this file except in compliance with the License. You may obtain +# a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT +# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the +# License for the specific language governing permissions and limitations +# under the License. +## + import asyncio import yaml import logging @@ -10,7 +26,8 @@ import sys import ROclient import ns import vim_sdn -from lcm_utils import versiontuple, LcmException, TaskRegistry +import netslice +from lcm_utils import versiontuple, LcmException, TaskRegistry, LcmExceptionExit # from osm_lcm import version as lcm_version, version_date as lcm_version_date, ROclient from osm_common import dbmemory, dbmongo, fslocal, msglocal, msgkafka @@ -23,16 +40,19 @@ from n2vc import version as n2vc_version __author__ = "Alfonso Tierno" -min_RO_version = [0, 5, 72] +min_RO_version = [0, 6, 0] min_n2vc_version = "0.0.2" -min_common_version = "0.1.7" +min_common_version = "0.1.11" # uncomment if LCM is installed as library and installed, and get them from __init__.py -lcm_version = '0.1.16' -lcm_version_date = '2018-10-10' +lcm_version = '0.1.31' +lcm_version_date = '2019-01-16' class Lcm: + ping_interval_pace = 120 # how many time ping is send once is confirmed all is running + ping_interval_boot = 5 # how many time ping is sent when booting + def __init__(self, config_file, loop=None): """ Init, Connect to database, filesystem storage, and messaging @@ -102,7 +122,7 @@ class Lcm: raise LcmException("Not compatible osm/N2VC version '{}'. Needed '{}' or higher".format( n2vc_version, min_n2vc_version)) # check version of common - if versiontuple(common_version) < versiontuple("0.1.7"): + if versiontuple(common_version) < versiontuple(min_common_version): raise LcmException("Not compatible osm/common version '{}'. Needed '{}' or higher".format( common_version, min_common_version)) @@ -139,7 +159,10 @@ class Lcm: raise LcmException(str(e)) self.ns = ns.NsLcm(self.db, self.msg, self.fs, self.lcm_tasks, self.ro_config, self.vca_config, self.loop) + self.netslice = netslice.NetsliceLcm(self.db, self.msg, self.fs, self.lcm_tasks, self.ro_config, + self.vca_config, self.loop) self.vim = vim_sdn.VimLcm(self.db, self.msg, self.fs, self.lcm_tasks, self.ro_config, self.loop) + self.wim = vim_sdn.WimLcm(self.db, self.msg, self.fs, self.lcm_tasks, self.ro_config, self.loop) self.sdn = vim_sdn.SdnLcm(self.db, self.msg, self.fs, self.lcm_tasks, self.ro_config, self.loop) async def check_RO_version(self): @@ -168,7 +191,7 @@ class Lcm: try: await self.msg.aiowrite("admin", "ping", {"from": "lcm", "to": "lcm"}, self.loop) # time between pings are low when it is not received and at starting - wait_time = 5 if not kafka_has_received else 120 + wait_time = self.ping_interval_boot if not kafka_has_received else self.ping_interval_pace if not self.pings_not_received: kafka_has_received = True self.pings_not_received += 1 @@ -190,138 +213,218 @@ class Lcm: wait_time = 1 if not first_start else 5 await asyncio.sleep(wait_time, loop=self.loop) + def kafka_read_callback(self, topic, command, params): + order_id = 1 + + if topic != "admin" and command != "ping": + self.logger.debug("Task kafka_read receives {} {}: {}".format(topic, command, params)) + self.consecutive_errors = 0 + self.first_start = False + order_id += 1 + if command == "exit": + raise LcmExceptionExit + elif command.startswith("#"): + return + elif command == "echo": + # just for test + print(params) + sys.stdout.flush() + return + elif command == "test": + asyncio.Task(self.test(params), loop=self.loop) + return + + if topic == "admin": + if command == "ping" and params["to"] == "lcm" and params["from"] == "lcm": + self.pings_not_received = 0 + return + elif topic == "ns": + if command == "instantiate": + # self.logger.debug("Deploying NS {}".format(nsr_id)) + nslcmop = params + nslcmop_id = nslcmop["_id"] + nsr_id = nslcmop["nsInstanceId"] + task = asyncio.ensure_future(self.ns.instantiate(nsr_id, nslcmop_id)) + self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_instantiate", task) + return + elif command == "terminate": + # self.logger.debug("Deleting NS {}".format(nsr_id)) + nslcmop = params + nslcmop_id = nslcmop["_id"] + nsr_id = nslcmop["nsInstanceId"] + self.lcm_tasks.cancel(topic, nsr_id) + task = asyncio.ensure_future(self.ns.terminate(nsr_id, nslcmop_id)) + self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_terminate", task) + return + elif command == "action": + # self.logger.debug("Update NS {}".format(nsr_id)) + nslcmop = params + nslcmop_id = nslcmop["_id"] + nsr_id = nslcmop["nsInstanceId"] + task = asyncio.ensure_future(self.ns.action(nsr_id, nslcmop_id)) + self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_action", task) + return + elif command == "scale": + # self.logger.debug("Update NS {}".format(nsr_id)) + nslcmop = params + nslcmop_id = nslcmop["_id"] + nsr_id = nslcmop["nsInstanceId"] + task = asyncio.ensure_future(self.ns.scale(nsr_id, nslcmop_id)) + self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_scale", task) + return + elif command == "show": + try: + db_nsr = self.db.get_one("nsrs", {"_id": nsr_id}) + print("nsr:\n _id={}\n operational-status: {}\n config-status: {}" + "\n detailed-status: {}\n deploy: {}\n tasks: {}" + "".format(nsr_id, db_nsr["operational-status"], db_nsr["config-status"], + db_nsr["detailed-status"], + db_nsr["_admin"]["deployed"], self.lcm_ns_tasks.get(nsr_id))) + except Exception as e: + print("nsr {} not found: {}".format(nsr_id, e)) + sys.stdout.flush() + return + elif command == "deleted": + return # TODO cleaning of task just in case should be done + elif command in ("terminated", "instantiated", "scaled", "actioned"): # "scaled-cooldown-time" + return + elif topic == "nsi": # netslice LCM processes (instantiate, terminate, etc) + if command == "instantiate": + # self.logger.debug("Instantiating Network Slice {}".format(nsilcmop["netsliceInstanceId"])) + nsilcmop = params + nsilcmop_id = nsilcmop["_id"] # slice operation id + nsir_id = nsilcmop["netsliceInstanceId"] # slice record id + task = asyncio.ensure_future(self.netslice.instantiate(nsir_id, nsilcmop_id)) + self.lcm_tasks.register("nsi", nsir_id, nsilcmop_id, "nsi_instantiate", task) + return + elif command == "terminate": + # self.logger.debug("Terminating Network Slice NS {}".format(nsilcmop["netsliceInstanceId"])) + nsilcmop = params + nsilcmop_id = nsilcmop["_id"] # slice operation id + nsir_id = nsilcmop["netsliceInstanceId"] # slice record id + self.lcm_tasks.cancel(topic, nsir_id) + task = asyncio.ensure_future(self.netslice.terminate(nsir_id, nsilcmop_id)) + self.lcm_tasks.register("nsi", nsir_id, nsilcmop_id, "nsi_terminate", task) + return + elif command == "show": + try: + db_nsir = self.db.get_one("nsirs", {"_id": nsir_id}) + print("nsir:\n _id={}\n operational-status: {}\n config-status: {}" + "\n detailed-status: {}\n deploy: {}\n tasks: {}" + "".format(nsir_id, db_nsir["operational-status"], db_nsir["config-status"], + db_nsir["detailed-status"], + db_nsir["_admin"]["deployed"], self.lcm_netslice_tasks.get(nsir_id))) + except Exception as e: + print("nsir {} not found: {}".format(nsir_id, e)) + sys.stdout.flush() + return + elif command == "deleted": + return # TODO cleaning of task just in case should be done + elif command in ("terminated", "instantiated", "scaled", "actioned"): # "scaled-cooldown-time" + return + elif topic == "vim_account": + vim_id = params["_id"] + if command == "create": + task = asyncio.ensure_future(self.vim.create(params, order_id)) + self.lcm_tasks.register("vim_account", vim_id, order_id, "vim_create", task) + return + elif command == "delete": + self.lcm_tasks.cancel(topic, vim_id) + task = asyncio.ensure_future(self.vim.delete(vim_id, order_id)) + self.lcm_tasks.register("vim_account", vim_id, order_id, "vim_delete", task) + return + elif command == "show": + print("not implemented show with vim_account") + sys.stdout.flush() + return + elif command == "edit": + task = asyncio.ensure_future(self.vim.edit(params, order_id)) + self.lcm_tasks.register("vim_account", vim_id, order_id, "vim_edit", task) + return + elif topic == "wim_account": + wim_id = params["_id"] + if command == "create": + task = asyncio.ensure_future(self.wim.create(params, order_id)) + self.lcm_tasks.register("wim_account", wim_id, order_id, "wim_create", task) + return + elif command == "delete": + self.lcm_tasks.cancel(topic, wim_id) + task = asyncio.ensure_future(self.wim.delete(wim_id, order_id)) + self.lcm_tasks.register("wim_account", wim_id, order_id, "wim_delete", task) + return + elif command == "show": + print("not implemented show with wim_account") + sys.stdout.flush() + return + elif command == "edit": + task = asyncio.ensure_future(self.wim.edit(params, order_id)) + self.lcm_tasks.register("wim_account", wim_id, order_id, "wim_edit", task) + return + elif topic == "sdn": + _sdn_id = params["_id"] + if command == "create": + task = asyncio.ensure_future(self.sdn.create(params, order_id)) + self.lcm_tasks.register("sdn", _sdn_id, order_id, "sdn_create", task) + return + elif command == "delete": + self.lcm_tasks.cancel(topic, _sdn_id) + task = asyncio.ensure_future(self.sdn.delete(_sdn_id, order_id)) + self.lcm_tasks.register("sdn", _sdn_id, order_id, "sdn_delete", task) + return + elif command == "edit": + task = asyncio.ensure_future(self.sdn.edit(params, order_id)) + self.lcm_tasks.register("sdn", _sdn_id, order_id, "sdn_edit", task) + return + self.logger.critical("unknown topic {} and command '{}'".format(topic, command)) + async def kafka_read(self): self.logger.debug("Task kafka_read Enter") - order_id = 1 # future = asyncio.Future() - consecutive_errors = 0 - first_start = True - while consecutive_errors < 10: + self.consecutive_errors = 0 + self.first_start = True + while self.consecutive_errors < 10: try: - topics = ("admin", "ns", "vim_account", "sdn") - topic, command, params = await self.msg.aioread(topics, self.loop) - if topic != "admin" and command != "ping": - self.logger.debug("Task kafka_read receives {} {}: {}".format(topic, command, params)) - consecutive_errors = 0 - first_start = False - order_id += 1 - if command == "exit": - print("Bye!") - break - elif command.startswith("#"): - continue - elif command == "echo": - # just for test - print(params) - sys.stdout.flush() - continue - elif command == "test": - asyncio.Task(self.test(params), loop=self.loop) - continue + topics = ("admin", "ns", "vim_account", "wim_account", "sdn", "nsi") + await self.msg.aioread(topics, self.loop, self.kafka_read_callback) - if topic == "admin": - if command == "ping" and params["to"] == "lcm" and params["from"] == "lcm": - self.pings_not_received = 0 - continue - elif topic == "ns": - if command == "instantiate": - # self.logger.debug("Deploying NS {}".format(nsr_id)) - nslcmop = params - nslcmop_id = nslcmop["_id"] - nsr_id = nslcmop["nsInstanceId"] - task = asyncio.ensure_future(self.ns.instantiate(nsr_id, nslcmop_id)) - self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_instantiate", task) - continue - elif command == "terminate": - # self.logger.debug("Deleting NS {}".format(nsr_id)) - nslcmop = params - nslcmop_id = nslcmop["_id"] - nsr_id = nslcmop["nsInstanceId"] - self.lcm_tasks.cancel(topic, nsr_id) - task = asyncio.ensure_future(self.ns.terminate(nsr_id, nslcmop_id)) - self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_terminate", task) - continue - elif command == "action": - # self.logger.debug("Update NS {}".format(nsr_id)) - nslcmop = params - nslcmop_id = nslcmop["_id"] - nsr_id = nslcmop["nsInstanceId"] - task = asyncio.ensure_future(self.ns.action(nsr_id, nslcmop_id)) - self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_action", task) - continue - elif command == "scale": - # self.logger.debug("Update NS {}".format(nsr_id)) - nslcmop = params - nslcmop_id = nslcmop["_id"] - nsr_id = nslcmop["nsInstanceId"] - task = asyncio.ensure_future(self.ns.scale(nsr_id, nslcmop_id)) - self.lcm_tasks.register("ns", nsr_id, nslcmop_id, "ns_scale", task) - continue - elif command == "show": - try: - db_nsr = self.db.get_one("nsrs", {"_id": nsr_id}) - print("nsr:\n _id={}\n operational-status: {}\n config-status: {}" - "\n detailed-status: {}\n deploy: {}\n tasks: {}" - "".format(nsr_id, db_nsr["operational-status"], db_nsr["config-status"], - db_nsr["detailed-status"], - db_nsr["_admin"]["deployed"], self.lcm_ns_tasks.get(nsr_id))) - except Exception as e: - print("nsr {} not found: {}".format(nsr_id, e)) - sys.stdout.flush() - continue - elif command == "deleted": - continue # TODO cleaning of task just in case should be done - elif command in ("terminated", "instantiated", "scaled", "actioned"): # "scaled-cooldown-time" - continue - elif topic == "vim_account": - vim_id = params["_id"] - if command == "create": - task = asyncio.ensure_future(self.vim.create(params, order_id)) - self.lcm_tasks.register("vim_account", vim_id, order_id, "vim_create", task) - continue - elif command == "delete": - self.lcm_tasks.cancel(topic, vim_id) - task = asyncio.ensure_future(self.vim.delete(vim_id, order_id)) - self.lcm_tasks.register("vim_account", vim_id, order_id, "vim_delete", task) - continue - elif command == "show": - print("not implemented show with vim_account") - sys.stdout.flush() - continue - elif command == "edit": - task = asyncio.ensure_future(self.vim.edit(params, order_id)) - self.lcm_tasks.register("vim_account", vim_id, order_id, "vim_edit", task) - continue - elif topic == "sdn": - _sdn_id = params["_id"] - if command == "create": - task = asyncio.ensure_future(self.sdn.create(params, order_id)) - self.lcm_tasks.register("sdn", _sdn_id, order_id, "sdn_create", task) - continue - elif command == "delete": - self.lcm_tasks.cancel(topic, _sdn_id) - task = asyncio.ensure_future(self.sdn.delete(_sdn_id, order_id)) - self.lcm_tasks.register("sdn", _sdn_id, order_id, "sdn_delete", task) - continue - elif command == "edit": - task = asyncio.ensure_future(self.sdn.edit(params, order_id)) - self.lcm_tasks.register("sdn", _sdn_id, order_id, "sdn_edit", task) - continue - self.logger.critical("unknown topic {} and command '{}'".format(topic, command)) + except LcmExceptionExit: + self.logger.debug("Bye!") + break except Exception as e: # if not first_start is the first time after starting. So leave more time and wait # to allow kafka starts - if consecutive_errors == 8 if not first_start else 30: + if self.consecutive_errors == 8 if not self.first_start else 30: self.logger.error("Task kafka_read task exit error too many errors. Exception: {}".format(e)) raise - consecutive_errors += 1 + self.consecutive_errors += 1 self.logger.error("Task kafka_read retrying after Exception {}".format(e)) - wait_time = 2 if not first_start else 5 + wait_time = 2 if not self.first_start else 5 await asyncio.sleep(wait_time, loop=self.loop) # self.logger.debug("Task kafka_read terminating") self.logger.debug("Task kafka_read exit") + def health_check(self): + + global exit_code + task = None + exit_code = 1 + + def health_check_callback(topic, command, params): + global exit_code + print("receiving callback {} {} {}".format(topic, command, params)) + if topic == "admin" and command == "ping" and params["to"] == "lcm" and params["from"] == "lcm": + # print("received LCM ping") + exit_code = 0 + task.cancel() + + try: + task = asyncio.ensure_future(self.msg.aioread(("admin",), self.loop, health_check_callback)) + self.loop.run_until_complete(task) + except Exception: + pass + exit(exit_code) + def start(self): # check RO version @@ -361,12 +464,14 @@ class Lcm: if not k.startswith("OSMLCM_"): continue k_items = k.lower().split("_") + if len(k_items) < 3: + continue + if k_items[1] in ("ro", "vca"): + # put in capital letter + k_items[1] = k_items[1].upper() c = conf try: for k_item in k_items[1:-1]: - if k_item in ("ro", "vca"): - # put in capital letter - k_item = k_item.upper() c = c[k_item] if k_items[-1] == "port": c[k_items[-1]] = int(v) @@ -384,6 +489,7 @@ class Lcm: def usage(): print("""Usage: {} [options] -c|--config [configuration_file]: loads the configuration file (default: ./nbi.cfg) + --health-check: do not run lcm, but inspect kafka bus to determine if lcm is healthy -h|--help: shows this help """.format(sys.argv[0])) # --log-socket-host HOST: send logs to this host") @@ -393,15 +499,18 @@ def usage(): if __name__ == '__main__': try: # load parameters and configuration - opts, args = getopt.getopt(sys.argv[1:], "hc:", ["config=", "help"]) + opts, args = getopt.getopt(sys.argv[1:], "hc:", ["config=", "help", "health-check"]) # TODO add "log-socket-host=", "log-socket-port=", "log-file=" config_file = None + health_check = None for o, a in opts: if o in ("-h", "--help"): usage() sys.exit() elif o in ("-c", "--config"): config_file = a + elif o == "--health-check": + health_check = True # elif o == "--log-socket-port": # log_socket_port = a # elif o == "--log-socket-host": @@ -412,17 +521,20 @@ if __name__ == '__main__': assert False, "Unhandled option" if config_file: if not path.isfile(config_file): - print("configuration file '{}' that not exist".format(config_file), file=sys.stderr) + print("configuration file '{}' not exist".format(config_file), file=sys.stderr) exit(1) else: for config_file in (__file__[:__file__.rfind(".")] + ".cfg", "./lcm.cfg", "/etc/osm/lcm.cfg"): if path.isfile(config_file): break else: - print("No configuration file 'nbi.cfg' found neither at local folder nor at /etc/osm/", file=sys.stderr) + print("No configuration file 'lcm.cfg' found neither at local folder nor at /etc/osm/", file=sys.stderr) exit(1) lcm = Lcm(config_file) - lcm.start() + if health_check: + lcm.health_check() + else: + lcm.start() except (LcmException, getopt.GetoptError) as e: print(str(e), file=sys.stderr) # usage()