1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
|
# Copyright (c) 2017-2018 Wind River Systems, Inc.
#
# 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 logging
import traceback
import json
from rest_framework import status
from rest_framework.response import Response
from rest_framework.views import APIView
from django.conf import settings
from common.msapi import extsys
from titanium_cloud.vesagent.tasks import scheduleBacklogs
from titanium_cloud.vesagent.event_domain.fault_vm import buildBacklog_fault_vm
from django.core.cache import cache
logger = logging.getLogger(__name__)
class VesAgentCtrl(APIView):
'''
control plane of VesAgent
Design tips:
1, vesagent are multiple processing workers
2, the runtime logic is simple: a vesagent worker polls the data source (vm/hypervisor/host/vim/etc.)
and then feeds the encoded data to VES.
3, the vesagent workers can be distributed to different clouds while latency/throughput is concerned,
this distributed deployment usually comes along with the distributed VES deployment.
So it is very likely that the collected data from different VIM/Cloud instance will be fed into
different VES endpoint, however, assuming that there will be at most one VES endpoint serving
any single VIM/Cloud instance.
4, According to VES specs, the collected data can be cataloged by domain:
domain : fault, heartbeat, measurementsForVfScaling, other, stateChange, syslog, thresholdCrossingAlert
As far as VIM/Cloud concerned, fault, heartbeat, measurementsForVfScaling, TCAalert are relevant.
5, the source of the collected data can be cataloged by eventSourceType:
eventSourceType: VNF/VNFC/VM
As far as VIM/Cloud concerned, only VM is relevant. This eventSourceType should be extended to cover
the data source of hypervisor, VIM, Host,Controller, PIM, etc.
6, the source of collected data should be specified explicitly,so is the domain of the collected data.
To specify the source: eventSourceType, uuid or name of the source
To specify the domain: domain
the specifications above will be provisioned as a vesagent backlog entry to a VIM/Cloud instance
to tell a vesagent worker that :
with regarding to that VIM/Cloud instance, what kind of data to be collected from which source .
7,the VES endpoint will be also specified for a VIM/Cloud instance, so that all collected data
will be fed into this VES endpoint
8, the vesagent backlog are stored into the respective cloud_region's property "cloud-extra-info",
which implies that those specifications can be CRUD either by ESR portal or the RestAPIs in this view, e.g.
"cloud-extra-info": {
...,
"vesagent_config":
{
"ves_subscription":{
"endpoint":"http://{VES IP}:{VES port}/{URI}",
"username":"{VES username}",
"password":"{VES password}",
},
"poll_interval_default" : "{default interval for polling}",
"backlogs":[
{
"domain":"fault"
"type":"vm",
"tenant":"{tenant name1}",
"source":"{VM name1}",
"poll_interval" : "{optional, interval for polling}",
},
{
"domain":"fault"
"type":"vm",
"tenant":"{tenant name2}",
"source":"{VM name2}",
"poll_interval" : "{optional, interval for polling}",
}
]
}
}
Idea: API dispatching to distributed M.C. service can be determined by Complex Object in AAI:
cloud-region has been assoicated to a Complex Object
M.C. plugin service instance should refer to the same Complex Object (by physical_locaton_id ?)
So the M.C. broker/API distributor/other approach will correlate the cloud-region with
corresponding M.C. plugin service instance.
Backlog built in cache:
maintain backlog in cache and VES agent workers
cache objects:
"VesAgentBacklogs.vimlist": [ list of vimid] ### will not expire forever
"VesAgentBacklogs.state.{vimdid}":
### will expire eventually to eliminate the garbage, expiration duration: 1hour?
{
"{backlog_uuid}": {
"timestamp": "{timestamp for last time of data collecting}",
"api_data": [list of data to populate the format string of the API link]
"last_event": {object, event reported to ves last time}"
}
}
"VesAgentBacklogs.config.{vimdid}": ### will not expire forever
{
"vimid": "{vim id}",
"subscription": {
"endpoint": "{ves endpoint, e.g. http://ves_ip:ves_port/eventListener/v5}",
"username": "{username}",
"password": "{password}"
}
"poll_interval_default" : "{default interval for polling}",
"backlogs":[
{
"backlog_uuid": "{uuid to identify the backlog}"
"domain":"fault"
"type":"vm",
"tenant":"{tenant name1}",
"source":"{VM name1}",
"poll_interval" : "{optional, interval in second for polling}",
"api_method": "{GET/POST/PUT/etc.}",
"api_link":"{API link to collect data, could be format string}",
"tenant_id": tenant_id,
"server_id": server_id,
},
{
"domain":"fault"
"type":"vm",
"tenant":"{tenant name2}",
"source":"{VM name2}",
"poll_interval" : "{optional, interval in second for polling}",
"api_method": "{GET/POST/PUT/etc.}",
"api_link":"{API link to collect data, could be format string}",
"tenant_id": tenant_id,
"server_id": server_id,
}
]
}
'''
def __init__(self):
self._logger = logger
self.proxy_prefix = settings.MULTICLOUD_PREFIX
def get(self, request, vimid=""):
'''
get blob of vesagent-config
:param request:
:param vimid:
:return:
'''
self._logger.info("vimid: %s" % vimid)
self._logger.debug("with META: %s" % request.META)
pass
def post(self, request, vimid=""):
'''
update the blob of vesagent-config, rebuild the backlog for the vesagent workers,
and start the vesagent workers if not started yet
Implication: the request to this API endpoint will build the backlog locally, hence only local VES agent workers
will process these backlogs, which conforms to distributed deployment of M.C. services which includes VES agents
:param request:{"vesagent_config":
{"ves_subscription":
{"endpoint":"http://127.0.0.1:9005/sample",
"username":"user","password":"password"},
"poll_interval_default":10,
"backlogs":
[
{"domain":"fault","type":"vm","tenant":"VIM","source":"onap-aaf"}
]
}
}
:param vimid:
:return:
'''
self._logger.info("vimid: %s" % vimid)
self._logger.debug("with META: %s, with data: %s" % (request.META, request.data))
try:
vesagent_config = None
if request.data is None or request.data.get("vesagent_config", None) is None:
#Try to load the vesagent_config out of cloud_region["cloud_extra_info"]
viminfo = extsys.get_vim_by_id(vimid)
cloud_extra_info_str = viminfo.get('cloud_extra_info', None)
cloud_extra_info = json.loads(cloud_extra_info_str) if cloud_extra_info_str is not None else None
vesagent_config = cloud_extra_info.get("vesagent_config", None) if cloud_extra_info is not None else None
else:
vesagent_config = request.data.get("vesagent_config", None)
if vesagent_config is None:
return Response(data={'vesagent_config is not provided'},
status=status.HTTP_400_BAD_REQUEST)
vesagent_backlogs = self.buildBacklogsOneVIM(vimid, vesagent_config)
# store back to cloud_extra_info
# tbd
except Exception as e:
self._logger.error("exception:%s" % str(e))
return Response(data={'error': str(e)},
status=status.HTTP_500_INTERNAL_SERVER_ERROR)
return Response(data={"vesagent_config":vesagent_config,
"vesagent_backlogs": vesagent_backlogs},
status=status.HTTP_201_CREATED)
def delete(self, request, vimid=""):
'''
delete the blob of vesagent-config, remove it from backlog and stop the vesagent worker if no backlog
:param request:
:param vimid:
:return:
'''
self._logger.info("vimid: %s" % vimid)
self._logger.debug("with META: %s" % request.META)
return Response(status=status.HTTP_200_OK)
def buildBacklogsOneVIM(self, vimid, vesagent_config = None):
'''
build and cache backlog for specific cloud region,spawn vesagent workers if needed
:param vimid:
:param vesagent_config: vesagent_config data in json object
:return:
'''
self._logger.info("vimid: %s" % vimid)
self._logger.debug("config data: %s" % vesagent_config)
VesAgentBacklogsConfig = None
try:
if vesagent_config :
# now rebuild the backlog
VesAgentBacklogsConfig = {
"vimid": vimid,
"poll_interval_default": vesagent_config.get("poll_interval_default", 0),
"subscription": vesagent_config.get("ves_subscription", None),
"backlogs": [self.buildBacklog(vimid, b) for b in vesagent_config.get("backlogs", [])]
}
# add/update the backlog into cache
VesAgentBacklogsConfigStr = json.dumps(VesAgentBacklogsConfig)
# cache forever
cache.set("VesAgentBacklogs.config.%s" % vimid, VesAgentBacklogsConfigStr, None)
# notify schduler
scheduleBacklogs.delay(vimid)
except Exception as e:
self._logger.error("exception:%s" % str(e))
VesAgentBacklogsConfig = {"error":"exception occurs during build backlogs"}
self._logger.info("return")
return VesAgentBacklogsConfig
def buildBacklog(self, vimid, backlog_input):
self._logger.info("build backlog for: %s" % vimid)
self._logger.debug("with input: %s" % backlog_input)
try:
if backlog_input["domain"] == "fault" and backlog_input["type"] == "vm":
return buildBacklog_fault_vm(vimid, backlog_input)
else:
self._logger.warn("return with failure: unsupported backlog domain:%s, type:%s"
% (backlog_input["domain"], backlog_input["type"] == "vm"))
return None
except Exception as e:
self._logger.error("exception:%s" % str(e))
return None
self._logger.info("return without backlog")
return None
|
