Coverage for k8sclient/k8sclient.py : 74%

Hot-keys on this page
r m x p toggle line displays
j k next/prev highlighted chunk
0 (zero) top of page
1 (one) first highlighted chunk
1# ============LICENSE_START=======================================================
2# org.onap.dcae
3# ================================================================================
4# Copyright (c) 2019-2020 AT&T Intellectual Property. All rights reserved.
5# Copyright (c) 2020 Pantheon.tech. All rights reserved.
6# Copyright (c) 2020-2021 Nokia. All rights reserved.
7# Copyright (c) 2020 J. F. Lucas. All rights reserved.
8# ================================================================================
9# Licensed under the Apache License, Version 2.0 (the "License");
10# you may not use this file except in compliance with the License.
11# You may obtain a copy of the License at
12#
13# http://www.apache.org/licenses/LICENSE-2.0
14#
15# Unless required by applicable law or agreed to in writing, software
16# distributed under the License is distributed on an "AS IS" BASIS,
17# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
18# See the License for the specific language governing permissions and
19# limitations under the License.
20# ============LICENSE_END=========================================================
21#
22import os
23import re
24import uuid
26from kubernetes import config, client, stream
28# Default values for readiness probe
29PROBE_DEFAULT_PERIOD = 15
30PROBE_DEFAULT_TIMEOUT = 1
32# Location of k8s cluster config file ("kubeconfig")
33K8S_CONFIG_PATH = "/opt/onap/kube/kubeconfig"
35# Regular expression for interval/timeout specification
36INTERVAL_SPEC = re.compile("^([0-9]+)(s|m|h)?$")
37# Conversion factors to seconds
38FACTORS = {None: 1, "s": 1, "m": 60, "h": 3600}
40# Regular expression for port mapping
41# group 1: container port
42# group 2: / + protocol
43# group 3: protocol
44# group 4: host port
45PORTS = re.compile("^([0-9]+)(/(udp|UDP|tcp|TCP))?:([0-9]+)$")
47# Constants for external_cert
48MOUNT_PATH = "/etc/onap/oom/certservice/certs/"
49KEYSTORE_PATH = MOUNT_PATH + "certServiceClient-keystore.jks"
50TRUSTSTORE_PATH = MOUNT_PATH + "truststore.jks"
51DEFAULT_CERT_TYPE = "p12"
54def _create_deployment_name(component_name):
55 return "dep-{0}".format(component_name)[:63]
58def _create_service_name(component_name):
59 return "{0}".format(component_name)[:63]
62def _create_exposed_service_name(component_name):
63 return ("x{0}".format(component_name))[:63]
66def _create_exposed_v6_service_name(component_name):
67 return ("x{0}-ipv6".format(component_name))[:63]
70def _configure_api(location=None):
71 # Look for a kubernetes config file
72 if os.path.exists(K8S_CONFIG_PATH):
73 config.load_kube_config(config_file=K8S_CONFIG_PATH, context=location, persist_config=False)
74 else:
75 # Maybe we're running in a k8s container and we can use info provided by k8s
76 # We would like to use:
77 # config.load_incluster_config()
78 # but this looks into os.environ for kubernetes host and port, and from
79 # the plugin those aren't visible. So we use the InClusterConfigLoader class,
80 # where we can set the environment to what we like.
81 # This is probably brittle! Maybe there's a better alternative.
82 localenv = {
83 config.incluster_config.SERVICE_HOST_ENV_NAME: "kubernetes.default.svc.cluster.local",
84 config.incluster_config.SERVICE_PORT_ENV_NAME: "443"
85 }
86 config.incluster_config.InClusterConfigLoader(
87 token_filename=config.incluster_config.SERVICE_TOKEN_FILENAME,
88 cert_filename=config.incluster_config.SERVICE_CERT_FILENAME,
89 environ=localenv
90 ).load_and_set()
93def _parse_interval(t):
94 """
95 Parse an interval specification
96 t can be
97 - a simple integer quantity, interpreted as seconds
98 - a string representation of a decimal integer, interpreted as seconds
99 - a string consisting of a represention of an decimal integer followed by a unit,
100 with "s" representing seconds, "m" representing minutes,
101 and "h" representing hours
102 Used for compatibility with the Docker plugin, where time intervals
103 for health checks were specified as strings with a number and a unit.
104 See 'intervalspec' above for the regular expression that's accepted.
105 """
106 m = INTERVAL_SPEC.match(str(t))
107 if m:
108 time = int(m.group(1)) * FACTORS[m.group(2)]
109 else:
110 raise ValueError("Bad interval specification: {0}".format(t))
111 return time
114def _create_probe(hc, port):
115 """ Create a Kubernetes probe based on info in the health check dictionary hc """
116 probe_type = hc['type']
117 probe = None
118 period = _parse_interval(hc.get('interval', PROBE_DEFAULT_PERIOD))
119 timeout = _parse_interval(hc.get('timeout', PROBE_DEFAULT_TIMEOUT))
120 if probe_type in ['http', 'https']:
121 probe = client.V1Probe(
122 failure_threshold=1,
123 initial_delay_seconds=5,
124 period_seconds=period,
125 timeout_seconds=timeout,
126 http_get=client.V1HTTPGetAction(
127 path=hc['endpoint'],
128 port=port,
129 scheme=probe_type.upper()
130 )
131 )
132 elif probe_type in ['script', 'docker']: 132 ↛ 142line 132 didn't jump to line 142, because the condition on line 132 was never false
133 probe = client.V1Probe(
134 failure_threshold=1,
135 initial_delay_seconds=5,
136 period_seconds=period,
137 timeout_seconds=timeout,
138 _exec=client.V1ExecAction(
139 command=hc['script'].split()
140 )
141 )
142 return probe
145def _create_resources(resources=None):
146 if resources is not None: 146 ↛ 153line 146 didn't jump to line 153, because the condition on line 146 was never false
147 resources_obj = client.V1ResourceRequirements(
148 limits=resources.get("limits"),
149 requests=resources.get("requests")
150 )
151 return resources_obj
152 else:
153 return None
156def _create_container_object(name, image, always_pull, **kwargs):
157 # Set up environment variables
158 # Copy any passed in environment variables
159 env = kwargs.get('env') or {}
160 env_vars = [client.V1EnvVar(name=k, value=env[k]) for k in env]
161 # Add POD_IP with the IP address of the pod running the container
162 pod_ip = client.V1EnvVarSource(field_ref=client.V1ObjectFieldSelector(field_path="status.podIP"))
163 env_vars.append(client.V1EnvVar(name="POD_IP", value_from=pod_ip))
165 # If a health check is specified, create a readiness/liveness probe
166 # (For an HTTP-based check, we assume it's at the first container port)
167 readiness = kwargs.get('readiness')
168 liveness = kwargs.get('liveness')
169 resources = kwargs.get('resources')
170 container_ports = kwargs.get('container_ports') or []
172 hc_port = container_ports[0][0] if container_ports else None
173 probe = _create_probe(readiness, hc_port) if readiness else None
174 live_probe = _create_probe(liveness, hc_port) if liveness else None
175 resources_obj = _create_resources(resources) if resources else None
176 port_objs = [client.V1ContainerPort(container_port=port, protocol=proto)
177 for port, proto in container_ports]
179 # Define container for pod
180 return client.V1Container(
181 name=name,
182 image=image,
183 image_pull_policy='Always' if always_pull else 'IfNotPresent',
184 env=env_vars,
185 ports=port_objs,
186 volume_mounts=kwargs.get('volume_mounts') or [],
187 resources=resources_obj,
188 readiness_probe=probe,
189 liveness_probe=live_probe
190 )
193def _create_deployment_object(component_name,
194 containers,
195 init_containers,
196 replicas,
197 volumes,
198 labels=None,
199 pull_secrets=None):
200 if labels is None: 200 ↛ 201line 200 didn't jump to line 201, because the condition on line 200 was never true
201 labels = {}
202 if pull_secrets is None: 202 ↛ 203line 202 didn't jump to line 203, because the condition on line 202 was never true
203 pull_secrets = []
204 deployment_name = _create_deployment_name(component_name)
206 # Label the pod with the deployment name, so we can find it easily
207 labels.update({"k8sdeployment": deployment_name})
209 # pull_secrets is a list of the names of the k8s secrets containing docker registry credentials
210 # See https://kubernetes.io/docs/concepts/containers/images/#specifying-imagepullsecrets-on-a-pod
211 ips = []
212 for secret in pull_secrets:
213 ips.append(client.V1LocalObjectReference(name=secret))
215 # Define pod template
216 template = client.V1PodTemplateSpec(
217 metadata=client.V1ObjectMeta(labels=labels),
218 spec=client.V1PodSpec(hostname=component_name,
219 containers=containers,
220 init_containers=init_containers,
221 volumes=volumes,
222 image_pull_secrets=ips)
223 )
225 # Define deployment spec
226 spec = client.V1DeploymentSpec(
227 replicas=replicas,
228 selector=client.V1LabelSelector(match_labels=labels),
229 template=template
230 )
232 # Create deployment object
233 deployment = client.V1Deployment(
234 api_version="apps/v1",
235 kind="Deployment",
236 metadata=client.V1ObjectMeta(name=deployment_name, labels=labels),
237 spec=spec
238 )
240 return deployment
243def _create_service_object(service_name, component_name, service_ports, annotations, labels, service_type, ip_family):
244 service_spec = client.V1ServiceSpec(
245 ports=service_ports,
246 selector={"app": component_name},
247 type=service_type,
248 ip_family=ip_family
249 )
250 if annotations: 250 ↛ 251line 250 didn't jump to line 251, because the condition on line 250 was never true
251 metadata = client.V1ObjectMeta(name=_create_service_name(service_name), labels=labels, annotations=annotations)
252 else:
253 metadata = client.V1ObjectMeta(name=_create_service_name(service_name), labels=labels)
255 service = client.V1Service(
256 kind="Service",
257 api_version="v1",
258 metadata=metadata,
259 spec=service_spec
260 )
261 return service
264def parse_ports(port_list):
265 """
266 Parse the port list into a list of container ports (needed to create the container)
267 and to a set of port mappings to set up k8s services.
268 """
269 container_ports = []
270 port_map = {}
271 for p in port_list:
272 ipv6 = False
273 if type(p) is dict:
274 ipv6 = "ipv6" in p and p['ipv6']
275 p = "".join(str(v) for v in p['concat'])
276 m = PORTS.match(p.strip())
277 if m:
278 cport = int(m.group(1))
279 hport = int(m.group(4))
280 if m.group(3):
281 proto = (m.group(3)).upper()
282 else:
283 proto = "TCP"
284 port = (cport, proto)
285 if port not in container_ports: 285 ↛ 287line 285 didn't jump to line 287, because the condition on line 285 was never false
286 container_ports.append(port)
287 port_map[(cport, proto, ipv6)] = hport
288 else:
289 raise ValueError("Bad port specification: {0}".format(p))
291 return container_ports, port_map
294def _parse_volumes(volume_list):
295 volumes = []
296 volume_mounts = []
297 for v in volume_list:
298 vname = str(uuid.uuid4())
299 vcontainer = v['container']['bind']
300 vro = (v['container'].get('mode') == 'ro')
301 if ('host' in v) and ('path' in v['host']):
302 vhost = v['host']['path']
303 volumes.append(client.V1Volume(name=vname, host_path=client.V1HostPathVolumeSource(path=vhost)))
304 if ('config_volume' in v) and ('name' in v['config_volume']):
305 vconfig_volume = v['config_volume']['name']
306 volumes.append(client.V1Volume(name=vname, config_map=client.V1ConfigMapVolumeSource(default_mode="0644",
307 name=vconfig_volume,
308 optional=True)))
309 volume_mounts.append(client.V1VolumeMount(name=vname, mount_path=vcontainer, read_only=vro))
311 return volumes, volume_mounts
314def _add_elk_logging_sidecar(containers, volumes, volume_mounts, component_name, log_info, filebeat):
315 if not log_info or not filebeat: 315 ↛ 316line 315 didn't jump to line 316, because the condition on line 315 was never true
316 return
317 log_dir = log_info.get("log_directory")
318 if not log_dir: 318 ↛ 319line 318 didn't jump to line 319, because the condition on line 318 was never true
319 return
320 sidecar_volume_mounts = []
322 # Create the volume for component log files and volume mounts for the component and sidecar containers
323 volumes.append(client.V1Volume(name="component-log", empty_dir=client.V1EmptyDirVolumeSource()))
324 volume_mounts.append(client.V1VolumeMount(name="component-log", mount_path=log_dir))
325 sc_path = log_info.get("alternate_fb_path") or "{0}/{1}".format(filebeat["log_path"], component_name)
326 sidecar_volume_mounts.append(client.V1VolumeMount(name="component-log", mount_path=sc_path))
328 # Create the volume for sidecar data and the volume mount for it
329 volumes.append(client.V1Volume(name="filebeat-data", empty_dir=client.V1EmptyDirVolumeSource()))
330 sidecar_volume_mounts.append(client.V1VolumeMount(name="filebeat-data", mount_path=filebeat["data_path"]))
332 # Create the volume for the sidecar configuration data and the volume mount for it
333 # The configuration data is in a k8s ConfigMap that should be created when DCAE is installed.
334 volumes.append(
335 client.V1Volume(name="filebeat-conf", config_map=client.V1ConfigMapVolumeSource(name=filebeat["config_map"])))
336 sidecar_volume_mounts.append(
337 client.V1VolumeMount(name="filebeat-conf", mount_path=filebeat["config_path"],
338 sub_path=filebeat["config_subpath"]))
340 # Finally create the container for the sidecar
341 containers.append(
342 _create_container_object("filebeat", filebeat["image"], False, volume_mounts=sidecar_volume_mounts))
345def _add_tls_init_container(ctx, init_containers, volumes, volume_mounts, tls_info, tls_config):
346 # Adds an InitContainer to the pod to set up TLS certificate information. For components that act as a server(
347 # tls_info["use_tls"] is True), the InitContainer will populate a directory with server and CA certificate
348 # materials in various formats. For other components (tls_info["use_tls"] is False, or tls_info is not
349 # specified), the InitContainer will populate a directory with CA certificate materials in PEM and JKS formats.
350 # In either case, the certificate directory is mounted onto the component container filesystem at the location
351 # specified by tls_info["component_cert_dir"], if present, otherwise at the configured default mount point (
352 # tls_config["component_cert_dir"]).
353 docker_image = tls_config["image"]
354 ctx.logger.info("Creating init container: TLS \n * [" + docker_image + "]")
356 cert_directory = tls_info.get("cert_directory") or tls_config.get("component_cert_dir")
357 env = {}
358 env["TLS_SERVER"] = "true" if tls_info.get("use_tls") else "false"
360 # Create the certificate volume and volume mounts
361 volumes.append(client.V1Volume(name="tls-info", empty_dir=client.V1EmptyDirVolumeSource()))
362 volume_mounts.append(client.V1VolumeMount(name="tls-info", mount_path=cert_directory))
363 init_volume_mounts = [client.V1VolumeMount(name="tls-info", mount_path=tls_config["cert_path"])]
365 # Create the init container
366 init_containers.append(
367 _create_container_object("init-tls", docker_image, False, volume_mounts=init_volume_mounts, env=env))
370def _add_external_tls_init_container(ctx, init_containers, volumes, external_cert, external_tls_config):
371 # Adds an InitContainer to the pod which will generate external TLS certificates.
372 docker_image = external_tls_config["image_tag"]
373 ctx.logger.info("Creating init container: external TLS \n * [" + docker_image + "]")
375 env = {}
376 output_path = external_cert.get("external_cert_directory")
377 if not output_path.endswith('/'): 377 ↛ 378line 377 didn't jump to line 378, because the condition on line 377 was never true
378 output_path += '/'
380 env["REQUEST_URL"] = external_tls_config.get("request_url")
381 env["REQUEST_TIMEOUT"] = external_tls_config.get("timeout")
382 env["OUTPUT_PATH"] = output_path + "external"
383 env["OUTPUT_TYPE"] = external_cert.get("cert_type")
384 env["CA_NAME"] = external_cert.get("ca_name")
385 env["COMMON_NAME"] = external_cert.get("external_certificate_parameters").get("common_name")
386 env["ORGANIZATION"] = external_tls_config.get("organization")
387 env["ORGANIZATION_UNIT"] = external_tls_config.get("organizational_unit")
388 env["LOCATION"] = external_tls_config.get("location")
389 env["STATE"] = external_tls_config.get("state")
390 env["COUNTRY"] = external_tls_config.get("country")
391 env["SANS"] = external_cert.get("external_certificate_parameters").get("sans")
392 env["KEYSTORE_PATH"] = KEYSTORE_PATH
393 env["KEYSTORE_PASSWORD"] = external_tls_config.get("keystore_password")
394 env["TRUSTSTORE_PATH"] = TRUSTSTORE_PATH
395 env["TRUSTSTORE_PASSWORD"] = external_tls_config.get("truststore_password")
397 # Create the volumes and volume mounts
398 sec = client.V1SecretVolumeSource(secret_name=external_tls_config.get("cert_secret_name"))
399 volumes.append(client.V1Volume(name="tls-volume", secret=sec))
400 init_volume_mounts = [
401 client.V1VolumeMount(name="tls-info", mount_path=external_cert.get("external_cert_directory")),
402 client.V1VolumeMount(name="tls-volume", mount_path=MOUNT_PATH)]
404 # Create the init container
405 init_containers.append(
406 _create_container_object("cert-service-client", docker_image, False, volume_mounts=init_volume_mounts, env=env))
409def _add_cert_post_processor_init_container(ctx, init_containers, tls_info, tls_config, external_cert,
410 cert_post_processor_config):
411 # Adds an InitContainer to the pod to merge TLS and external TLS truststore into single file.
412 docker_image = cert_post_processor_config["image_tag"]
413 ctx.logger.info("Creating init container: cert post processor \n * [" + docker_image + "]")
415 tls_cert_dir = tls_info.get("cert_directory") or tls_config.get("component_cert_dir")
416 if not tls_cert_dir.endswith('/'): 416 ↛ 419line 416 didn't jump to line 419, because the condition on line 416 was never false
417 tls_cert_dir += '/'
419 tls_cert_file_path = tls_cert_dir + "trust.jks"
420 tls_cert_file_pass = tls_cert_dir + "trust.pass"
422 ext_cert_dir = tls_cert_dir + "external/"
424 output_type = (external_cert.get("cert_type") or DEFAULT_CERT_TYPE).lower()
425 ext_truststore_path = ext_cert_dir + "truststore." + _get_file_extension(output_type)
426 ext_truststore_pass = ''
427 if output_type != 'pem': 427 ↛ 430line 427 didn't jump to line 430, because the condition on line 427 was never false
428 ext_truststore_pass = ext_cert_dir + "truststore.pass"
430 env = {"TRUSTSTORES_PATHS": tls_cert_file_path + ":" + ext_truststore_path,
431 "TRUSTSTORES_PASSWORDS_PATHS": tls_cert_file_pass + ":" + ext_truststore_pass,
432 "KEYSTORE_SOURCE_PATHS": _get_keystore_source_paths(output_type, ext_cert_dir),
433 "KEYSTORE_DESTINATION_PATHS": _get_keystore_destination_paths(output_type, tls_cert_dir)}
435 ctx.logger.info("TRUSTSTORES_PATHS: " + env["TRUSTSTORES_PATHS"])
436 ctx.logger.info("TRUSTSTORES_PASSWORDS_PATHS: " + env["TRUSTSTORES_PASSWORDS_PATHS"])
437 ctx.logger.info("KEYSTORE_SOURCE_PATHS: " + env["KEYSTORE_SOURCE_PATHS"])
438 ctx.logger.info("KEYSTORE_DESTINATION_PATHS: " + env["KEYSTORE_DESTINATION_PATHS"])
440 # Create the volumes and volume mounts
441 init_volume_mounts = [client.V1VolumeMount(name="tls-info", mount_path=tls_cert_dir)]
443 # Create the init container
444 init_containers.append(
445 _create_container_object("cert-post-processor", docker_image, False, volume_mounts=init_volume_mounts, env=env))
448def _get_file_extension(output_type):
449 return {
450 'p12': 'p12',
451 'pem': 'pem',
452 'jks': 'jks',
453 }[output_type]
456def _get_keystore_source_paths(output_type, ext_cert_dir):
457 source_paths_template = {
458 'p12': "{0}keystore.p12:{0}keystore.pass",
459 'jks': "{0}keystore.jks:{0}keystore.pass",
460 'pem': "{0}keystore.pem:{0}key.pem",
461 }[output_type]
462 return source_paths_template.format(ext_cert_dir)
465def _get_keystore_destination_paths(output_type, tls_cert_dir):
466 destination_paths_template = {
467 'p12': "{0}cert.p12:{0}p12.pass",
468 'jks': "{0}cert.jks:{0}jks.pass",
469 'pem': "{0}cert.pem:{0}key.pem",
470 }[output_type]
471 return destination_paths_template.format(tls_cert_dir)
474def _process_port_map(port_map):
475 service_ports = [] # Ports exposed internally on the k8s network
476 exposed_ports = [] # Ports to be mapped to ports on the k8s nodes via NodePort
477 exposed_ports_ipv6 = []
478 for (cport, proto, ipv6), hport in port_map.items():
479 name = "xport-{0}-{1}".format(proto[0].lower(), cport)
480 cport = int(cport)
481 hport = int(hport)
482 port = client.V1ServicePort(port=cport, protocol=proto, name=name[1:])
483 if port not in service_ports: 483 ↛ 485line 483 didn't jump to line 485, because the condition on line 483 was never false
484 service_ports.append(port)
485 if hport != 0: 485 ↛ 486line 485 didn't jump to line 486, because the condition on line 485 was never true
486 if ipv6:
487 exposed_ports_ipv6.append(client.V1ServicePort(port=cport, protocol=proto, node_port=hport, name=name))
488 else:
489 exposed_ports.append(client.V1ServicePort(port=cport, protocol=proto, node_port=hport, name=name))
490 return service_ports, exposed_ports, exposed_ports_ipv6
493def _service_exists(location, namespace, component_name):
494 exists = False
495 try:
496 _configure_api(location)
497 client.CoreV1Api().read_namespaced_service(_create_service_name(component_name), namespace)
498 exists = True
499 except client.rest.ApiException:
500 pass
502 return exists
505def _patch_deployment(location, namespace, deployment, modify):
506 '''
507 Gets the current spec for 'deployment' in 'namespace'
508 in the k8s cluster at 'location',
509 uses the 'modify' function to change the spec,
510 then sends the updated spec to k8s.
511 '''
512 _configure_api(location)
514 # Get deployment spec
515 spec = client.AppsV1Api().read_namespaced_deployment(deployment, namespace)
517 # Apply changes to spec
518 spec = modify(spec)
520 # Patch the deploy with updated spec
521 client.AppsV1Api().patch_namespaced_deployment(deployment, namespace, spec)
524def _execute_command_in_pod(location, namespace, pod_name, command):
525 '''
526 Execute the command (specified by an argv-style list in the "command" parameter) in
527 the specified pod in the specified namespace at the specified location.
528 For now at least, we use this only to
529 run a notification script in a pod after a configuration change.
531 The straightforward way to do this is with the V1 Core API function "connect_get_namespaced_pod_exec".
532 Unfortunately due to a bug/limitation in the Python client library, we can't call it directly.
533 We have to make the API call through a Websocket connection using the kubernetes.stream wrapper API.
534 I'm following the example code at https://github.com/kubernetes-client/python/blob/master/examples/exec.py.
535 There are several issues tracking this, in various states. It isn't clear that there will ever
536 be a fix.
537 - https://github.com/kubernetes-client/python/issues/58
538 - https://github.com/kubernetes-client/python/issues/409
539 - https://github.com/kubernetes-client/python/issues/526
541 The main consequence of the workaround using "stream" is that the caller does not get an indication
542 of the exit code returned by the command when it completes execution. It turns out that the
543 original implementation of notification in the Docker plugin did not use this result, so we can
544 still match the original notification functionality.
546 The "stream" approach returns a string containing any output sent by the command to stdout or stderr.
547 We'll return that so it can logged.
548 '''
549 _configure_api(location)
550 try:
551 output = stream.stream(client.CoreV1Api().connect_get_namespaced_pod_exec,
552 name=pod_name,
553 namespace=namespace,
554 command=command,
555 stdout=True,
556 stderr=True,
557 stdin=False,
558 tty=False)
559 except client.rest.ApiException as e:
560 # If the exception indicates the pod wasn't found, it's not a fatal error.
561 # It existed when we enumerated the pods for the deployment but no longer exists.
562 # Unfortunately, the only way to distinguish a pod not found from any other error
563 # is by looking at the reason text.
564 # (The ApiException's "status" field should contain the HTTP status code, which would
565 # be 404 if the pod isn't found, but empirical testing reveals that "status" is set
566 # to zero.)
567 if "404 not found" in e.reason.lower():
568 output = "Pod not found"
569 else:
570 raise e
572 return {"pod": pod_name, "output": output}
575def deploy(ctx, namespace, component_name, image, replicas, always_pull, k8sconfig, **kwargs):
576 """
577 This will create a k8s Deployment and, if needed, one or two k8s Services.
578 (We are being opinionated in our use of k8s... this code decides what k8s abstractions and features to use.
579 We're not exposing k8s to the component developer and the blueprint author.
580 This is a conscious choice. We want to use k8s in a controlled, consistent way, and we want to hide
581 the details from the component developer and the blueprint author.)
583 namespace: the Kubernetes namespace into which the component is deployed
584 component_name: the component name, used to derive names of Kubernetes entities
585 image: the docker image for the component being deployed
586 replica: the number of instances of the component to be deployed
587 always_pull: boolean flag, indicating that Kubernetes should always pull a new copy of
588 the Docker image for the component, even if it is already present on the Kubernetes node.
589 k8sconfig contains:
590 - image_pull_secrets: a list of names of image pull secrets that can be used for retrieving images.
591 (DON'T PANIC: these are just the names of secrets held in the Kubernetes secret store.)
592 - filebeat: a dictionary of filebeat sidecar parameters:
593 "log_path" : mount point for log volume in filebeat container
594 "data_path" : mount point for data volume in filebeat container
595 "config_path" : mount point for config volume in filebeat container
596 "config_subpath" : subpath for config data in filebeat container
597 "config_map" : ConfigMap holding the filebeat configuration
598 "image": Docker image to use for filebeat
599 - tls: a dictionary of TLS-related information:
600 "cert_path": mount point for certificate volume in init container
601 "image": Docker image to use for TLS init container
602 "component_cert_dir" : default mount point for certs
603 - cert_post_processor: a dictionary of cert_post_processor information:
604 "image_tag": docker image to use for cert-post-processor init container
605 kwargs may have:
606 - volumes: array of volume objects, where a volume object is:
607 {"host":{"path": "/path/on/host"}, "container":{"bind":"/path/on/container","mode":"rw_or_ro"}
608 - ports: array of strings in the form "container_port:host_port"
609 - env: map of name-value pairs ( {name0: value0, name1: value1...}
610 - log_info: an object with info for setting up ELK logging, with the form:
611 {"log_directory": "/path/to/container/log/directory", "alternate_fb_path" : "/alternate/sidecar/log/path"}
612 - tls_info: an object with info for setting up TLS (HTTPS), with the form:
613 {"use_tls": true, "cert_directory": "/path/to/container/cert/directory" }
614 - external_cert: an object with information for setting up the init container for external certificates creation, with the form:
615 {"external_cert":
616 "external_cert_directory": "/path/to/directory_where_certs_should_be_placed",
617 "use_external_tls": true or false,
618 "ca_name": "ca-name-value",
619 "cert_type": "P12" or "JKS" or "PEM",
620 "external_certificate_parameters":
621 "common_name": "common-name-value",
622 "sans": "sans-value"}
623 - labels: dict with label-name/label-value pairs, e.g. {"cfydeployment" : "lsdfkladflksdfsjkl", "cfynode":"mycomponent"}
624 These label will be set on all the pods deployed as a result of this deploy() invocation.
625 - resources: dict with optional "limits" and "requests" resource requirements, each a dict containing:
626 - cpu: number CPU usage, like 0.5
627 - memory: string memory requirement, like "2Gi"
628 - readiness: dict with health check info; if present, used to create a readiness probe for the main container. Includes:
629 - type: check is done by making http(s) request to an endpoint ("http", "https") or by exec'ing a script in the container ("script", "docker")
630 - interval: period (in seconds) between probes
631 - timeout: time (in seconds) to allow a probe to complete
632 - endpoint: the path portion of the URL that points to the readiness endpoint for "http" and "https" types
633 - path: the full path to the script to be executed in the container for "script" and "docker" types
634 - liveness: dict with health check info; if present, used to create a liveness probe for the main container. Includes:
635 - type: check is done by making http(s) request to an endpoint ("http", "https") or by exec'ing a script in the container ("script", "docker")
636 - interval: period (in seconds) between probes
637 - timeout: time (in seconds) to allow a probe to complete
638 - endpoint: the path portion of the URL that points to the liveness endpoint for "http" and "https" types
639 - path: the full path to the script to be executed in the container for "script" and "docker" types
640 - k8s_location: name of the Kubernetes location (cluster) where the component is to be deployed
642 """
644 deployment_ok = False
645 cip_service_created = False
646 deployment_description = {
647 "namespace": namespace,
648 "location": kwargs.get("k8s_location"),
649 "deployment": '',
650 "services": []
651 }
653 try:
655 # Get API handles
656 _configure_api(kwargs.get("k8s_location"))
657 core = client.CoreV1Api()
658 k8s_apps_v1_api_client = client.AppsV1Api()
660 # Parse the port mapping
661 container_ports, port_map = parse_ports(kwargs.get("ports", []))
663 # Parse the volumes list into volumes and volume_mounts for the deployment
664 volumes, volume_mounts = _parse_volumes(kwargs.get("volumes", []))
666 # Initialize the list of containers that will be part of the pod
667 containers = []
668 init_containers = []
670 # Set up the ELK logging sidecar container, if needed
671 _add_elk_logging_sidecar(containers, volumes, volume_mounts, component_name, kwargs.get("log_info"),
672 k8sconfig.get("filebeat"))
674 # Set up TLS information
675 _add_tls_init_container(ctx, init_containers, volumes, volume_mounts, kwargs.get("tls_info") or {},
676 k8sconfig.get("tls"))
678 # Set up external TLS information
679 external_cert = kwargs.get("external_cert")
680 if external_cert and external_cert.get("use_external_tls"):
681 _add_external_tls_init_container(ctx, init_containers, volumes, external_cert,
682 k8sconfig.get("external_cert"))
683 _add_cert_post_processor_init_container(ctx, init_containers, kwargs.get("tls_info") or {},
684 k8sconfig.get("tls"), external_cert,
685 k8sconfig.get("cert_post_processor"))
687 # Create the container for the component
688 # Make it the first container in the pod
689 container_args = {key: kwargs.get(key) for key in ("env", "readiness", "liveness", "resources")}
690 container_args['container_ports'] = container_ports
691 container_args['volume_mounts'] = volume_mounts
692 containers.insert(0, _create_container_object(component_name, image, always_pull, **container_args))
694 # Build the k8s Deployment object
695 labels = kwargs.get("labels", {})
696 labels["app"] = component_name
697 dep = _create_deployment_object(component_name, containers, init_containers, replicas, volumes, labels,
698 pull_secrets=k8sconfig["image_pull_secrets"])
700 # Have k8s deploy it
701 k8s_apps_v1_api_client.create_namespaced_deployment(namespace, dep)
702 deployment_ok = True
703 deployment_description["deployment"] = _create_deployment_name(component_name)
705 # Create service(s), if a port mapping is specified
706 if port_map: 706 ↛ 741line 706 didn't jump to line 741, because the condition on line 706 was never false
707 service_ports, exposed_ports, exposed_ports_ipv6 = _process_port_map(port_map)
709 # Create a ClusterIP service for access via the k8s network
710 service = _create_service_object(_create_service_name(component_name), component_name, service_ports, None,
711 labels, "ClusterIP", "IPv4")
712 core.create_namespaced_service(namespace, service)
713 cip_service_created = True
714 deployment_description["services"].append(_create_service_name(component_name))
716 # If there are ports to be exposed on the k8s nodes, create a "NodePort" service
717 if exposed_ports: 717 ↛ 718line 717 didn't jump to line 718
718 exposed_service = \
719 _create_service_object(_create_exposed_service_name(component_name), component_name, exposed_ports,
720 '', labels, "NodePort", "IPv4")
721 core.create_namespaced_service(namespace, exposed_service)
722 deployment_description["services"].append(_create_exposed_service_name(component_name))
724 if exposed_ports_ipv6: 724 ↛ 725line 724 didn't jump to line 725
725 exposed_service_ipv6 = \
726 _create_service_object(_create_exposed_v6_service_name(component_name), component_name,
727 exposed_ports_ipv6, '', labels, "NodePort", "IPv6")
728 core.create_namespaced_service(namespace, exposed_service_ipv6)
729 deployment_description["services"].append(_create_exposed_v6_service_name(component_name))
731 except Exception as e:
732 # If the ClusterIP service was created, delete the service:
733 if cip_service_created:
734 core.delete_namespaced_service(_create_service_name(component_name), namespace)
735 # If the deployment was created but not the service, delete the deployment
736 if deployment_ok:
737 client.AppsV1Api().delete_namespaced_deployment(_create_deployment_name(component_name), namespace,
738 body=client.V1DeleteOptions())
739 raise e
741 return dep, deployment_description
744def undeploy(deployment_description):
745 _configure_api(deployment_description["location"])
747 namespace = deployment_description["namespace"]
749 # remove any services associated with the component
750 for service in deployment_description["services"]:
751 client.CoreV1Api().delete_namespaced_service(service, namespace)
753 # Have k8s delete the underlying pods and replicaset when deleting the deployment.
754 options = client.V1DeleteOptions(propagation_policy="Foreground")
755 client.AppsV1Api().delete_namespaced_deployment(deployment_description["deployment"], namespace, body=options)
758def is_available(location, namespace, component_name):
759 _configure_api(location)
760 dep_status = client.AppsV1Api().read_namespaced_deployment_status(_create_deployment_name(component_name),
761 namespace)
762 # Check if the number of available replicas is equal to the number requested and that the replicas match the
763 # current spec This check can be used to verify completion of an initial deployment, a scale operation,
764 # or an update operation
765 return dep_status.status.available_replicas == dep_status.spec.replicas and dep_status.status.updated_replicas == dep_status.spec.replicas
768def scale(deployment_description, replicas):
769 """ Trigger a scaling operation by updating the replica count for the Deployment """
771 def update_replica_count(spec):
772 spec.spec.replicas = replicas
773 return spec
775 _patch_deployment(deployment_description["location"], deployment_description["namespace"],
776 deployment_description["deployment"], update_replica_count)
779def upgrade(deployment_description, image, container_index=0):
780 """ Trigger a rolling upgrade by sending a new image name/tag to k8s """
782 def update_image(spec):
783 spec.spec.template.spec.containers[container_index].image = image
784 return spec
786 _patch_deployment(deployment_description["location"], deployment_description["namespace"],
787 deployment_description["deployment"], update_image)
790def rollback(deployment_description, rollback_to=0):
791 """
792 Undo upgrade by rolling back to a previous revision of the deployment.
793 By default, go back one revision.
794 rollback_to can be used to supply a specific revision number.
795 Returns the image for the app container and the replica count from the rolled-back deployment
796 """
797 '''
798 2018-07-13
799 Currently this does not work due to a bug in the create_namespaced_deployment_rollback() method.
800 The k8s python client code throws an exception while processing the response from the API.
801 See:
802 - https://github.com/kubernetes-client/python/issues/491
803 - https://github.com/kubernetes/kubernetes/pull/63837
804 The fix has been merged into the master branch but is not in the latest release.
805 '''
806 _configure_api(deployment_description["location"])
807 deployment = deployment_description["deployment"]
808 namespace = deployment_description["namespace"]
810 # Initiate the rollback
811 client.ExtensionsV1beta1Api().create_namespaced_deployment_rollback(
812 deployment,
813 namespace,
814 client.AppsV1beta1DeploymentRollback(name=deployment,
815 rollback_to=client.AppsV1beta1RollbackConfig(revision=rollback_to)))
817 # Read back the spec for the rolled-back deployment
818 spec = client.AppsV1Api().read_namespaced_deployment(deployment, namespace)
819 return spec.spec.template.spec.containers[0].image, spec.spec.replicas
822def execute_command_in_deployment(deployment_description, command):
823 """
824 Enumerates the pods in the k8s deployment identified by "deployment_description",
825 then executes the command (represented as an argv-style list) in "command" in
826 container 0 (the main application container) each of those pods.
828 Note that the sets of pods associated with a deployment can change over time. The
829 enumeration is a snapshot at one point in time. The command will not be executed in
830 pods that are created after the initial enumeration. If a pod disappears after the
831 initial enumeration and before the command is executed, the attempt to execute the
832 command will fail. This is not treated as a fatal error.
834 This approach is reasonable for the one current use case for "execute_command": running a
835 script to notify a container that its configuration has changed as a result of a
836 policy change. In this use case, the new configuration information is stored into
837 the configuration store (Consul), the pods are enumerated, and the command is executed.
838 If a pod disappears after the enumeration, the fact that the command cannot be run
839 doesn't matter--a nonexistent pod doesn't need to be reconfigured. Similarly, a pod that
840 comes up after the enumeration will get its initial configuration from the updated version
841 in Consul.
843 The optimal solution here would be for k8s to provide an API call to execute a command in
844 all of the pods for a deployment. Unfortunately, k8s does not provide such a call--the
845 only call provided by k8s operates at the pod level, not the deployment level.
847 Another interesting k8s factoid: there's no direct way to list the pods belong to a
848 particular k8s deployment. The deployment code above sets a label ("k8sdeployment") on
849 the pod that has the k8s deployment name. To list the pods, the code below queries for
850 pods with the label carrying the deployment name.
851 """
852 location = deployment_description["location"]
853 _configure_api(location)
854 deployment = deployment_description["deployment"]
855 namespace = deployment_description["namespace"]
857 # Get names of all the running pods belonging to the deployment
858 pod_names = [pod.metadata.name for pod in client.CoreV1Api().list_namespaced_pod(
859 namespace=namespace,
860 label_selector="k8sdeployment={0}".format(deployment),
861 field_selector="status.phase=Running"
862 ).items]
864 # Execute command in the running pods
865 return [_execute_command_in_pod(location, namespace, pod_name, command)
866 for pod_name in pod_names]