ddf28104a3
We've been using old cirros-0.4.0 for long time, and don't need to fix the version of cirros to 0.4.0. This patch is update to the latest 0.5.2. Most of the changes in this patch is replacing `image`, `hash` and `version` in TOSCA definition files. The rest of them are: * Replace images. First one is only an actual image, and rest of them are empty text files which is a trick for reducing total size in tacker repo. So, text files are not replaced, but renamed simply. * tacker/tests/etc/samples/etsi/nfv/common/Files/images/cirros-0.5.2-x86_64-disk.img * samples/vnf_packages/Files/images/cirros-0.5.2-x86_64-disk.img * tacker/tests/etc/samples/etsi/nfv/test_csar_utils_data/sample_vnfpkg_no_meta_single_vnfd/Files/images/cirros-0.5.2-x86_64-disk.img * tacker/tests/etc/samples/sample_vnf_package_csar_in_meta_and_manifest/Files/images/cirros-0.5.2-x86_64-disk.img * Update functional tests verifying images. A refactor of defining package size is included. * Update devstack script, and helper scripts under `contrib` dir. * Update API references and documents describing usage of the image. Signed-off-by: Yasufumi Ogawa <yasufum.o@gmail.com> Implements: blueprint update-cirros-052 Change-Id: I081b791dc986ab10f4845784b3e19dbb626e3cac
13 KiB
Alarm monitoring framework
This document describes how to use alarm-based monitoring driver in Tacker.
Sample TOSCA with monitoring policy
The following example shows monitoring policy using TOSCA template. The target (VDU1) of the monitoring policy in this example need to be described firstly like other TOSCA templates in Tacker.
policies:
- vdu1_cpu_usage_monitoring_policy:
type: tosca.policies.tacker.Alarming
triggers:
vdu_hcpu_usage_respawning:
event_type:
type: tosca.events.resource.utilization
implementation: ceilometer
metric: cpu_util
condition:
threshold: 50
constraint: utilization greater_than 50%
granularity: 600
evaluations: 1
aggregation_method: mean
resource_type: instance
comparison_operator: gt
metadata: VDU1
action: [respawn]Alarm framework already supported the some default backend actions like scaling, respawn, log, and log_and_kill.
Tacker users could change the desired action as described in the above example. Until now, the backend actions could be pointed to the specific policy which is also described in TOSCA template like scaling policy. The integration between alarming monitoring and scaling was also supported by Alarm monitor in Tacker:
tosca_definitions_version: tosca_simple_profile_for_nfv_1_0_0
description: Demo example
metadata:
template_name: sample-tosca-vnfd
topology_template:
node_templates:
VDU1:
type: tosca.nodes.nfv.VDU.Tacker
capabilities:
nfv_compute:
properties:
disk_size: 1 GB
mem_size: 512 MB
num_cpus: 2
properties:
image: cirros-0.5.2-x86_64-disk
mgmt_driver: noop
availability_zone: nova
metadata: {metering.server_group: SG1}
CP1:
type: tosca.nodes.nfv.CP.Tacker
properties:
management: true
anti_spoofing_protection: false
requirements:
- virtualLink:
node: VL1
- virtualBinding:
node: VDU1
VL1:
type: tosca.nodes.nfv.VL
properties:
network_name: net_mgmt
vendor: Tacker
policies:
- SP1:
type: tosca.policies.tacker.Scaling
targets: [VDU1]
properties:
increment: 1
cooldown: 120
min_instances: 1
max_instances: 3
default_instances: 1
- vdu_cpu_usage_monitoring_policy:
type: tosca.policies.tacker.Alarming
triggers:
vdu_hcpu_usage_scaling_out:
event_type:
type: tosca.events.resource.utilization
implementation: ceilometer
metric: cpu_util
condition:
threshold: 80
constraint: utilization greater_than 80%
granularity: 300
evaluations: 1
aggregation_method: mean
resource_type: instance
comparison_operator: gt
metadata: SG1
action: [SP1]
vdu_lcpu_usage_scaling_in:
event_type:
type: tosca.events.resource.utilization
implementation: ceilometer
metric: cpu_util
condition:
threshold: 10
constraint: utilization less_than 10%
granularity: 300
evaluations: 1
aggregation_method: mean
resource_type: instance
comparison_operator: lt
metadata: SG1
action: [SP1]NOTE: metadata defined in VDU properties must be matched with metadata in monitoring policy
How to setup environment
If OpenStack Devstack is used to test alarm monitoring in Tacker, OpenStack Ceilometer and Aodh plugins will need to be enabled in local.conf:
enable_plugin ceilometer https://opendev.org/openstack/ceilometer master
enable_plugin aodh https://opendev.org/openstack/aodh master
How to monitor VNFs via alarm triggers
How to setup alarm configuration
Tacker provides templates that implemented Ceilometer as alarm for monitoring VNFs, which are located in tacker/samples/tosca-templates/vnfd.
- tosca-vnfd-alarm-multi-actions.yaml
- tosca-vnfd-alarm-respawn.yaml
- tosca-vnfd-alarm-scale.yaml
The following commands shows creating VNF with alarms for scaling in and out.
$ cd ~/tacker/samples/tosca-templates/vnfd
$ openstack vnf create --vnfd-template tosca-vnfd-alarm-scale.yaml VNF1Firstly, vnfd and vnf need to be created successfully using pre-defined TOSCA template for alarm monitoring. Then, in order to know whether alarm configuration defined in Tacker is successfully passed to Ceilometer, Tacker users could use CLI:
$ openstack alarm list
+--------------------------------------+--------------------------------------------+-----------------------------------------------------------------------------------+-------------------+----------+---------+
| alarm_id | type | name | state | severity | enabled |
+--------------------------------------+--------------------------------------------+-----------------------------------------------------------------------------------+-------------------+----------+---------+
| f418ebf8-f8a6-4991-8f0d-938e38434411 | gnocchi_aggregation_by_resources_threshold | VNF1_7582cdf4-58ed-4df8-8fa2-c15938adf70b-vdu_hcpu_usage_scaling_out-4imzw3c7cicb | insufficient data | low | True |
| 70d86622-940a-4bc3-87c2-d5dfbb01bbea | gnocchi_aggregation_by_resources_threshold | VNF1_7582cdf4-58ed-4df8-8fa2-c15938adf70b-vdu_lcpu_usage_scaling_in-dwvdvbegiqdk | insufficient data | low | True |
+--------------------------------------+--------------------------------------------+-----------------------------------------------------------------------------------+-------------------+----------+---------+$ openstack alarm show 70d86622-940a-4bc3-87c2-d5dfbb01bbea
+---------------------------+------------------------------------------------------------------------------------------------------------------+
| Field | Value |
+---------------------------+------------------------------------------------------------------------------------------------------------------+
| aggregation_method | mean |
| alarm_actions | [u'http://ubuntu:9890/v1.0/vnfs/7582cdf4-58ed-4df8-8fa2-c15938adf70b/vdu_lcpu_usage_scaling_in/SP1-in/v2fq7rd7'] |
| alarm_id | 70d86622-940a-4bc3-87c2-d5dfbb01bbea |
| comparison_operator | lt |
| description | utilization less_than 10% |
| enabled | True |
| evaluation_periods | 1 |
| granularity | 60 |
| insufficient_data_actions | [] |
| metric | cpu_util |
| name | VNF1_7582cdf4-58ed-4df8-8fa2-c15938adf70b-vdu_lcpu_usage_scaling_in-dwvdvbegiqdk |
| ok_actions | [] |
| project_id | b5e054a3861b4da2b084aca9530096be |
| query | {"=": {"server_group": "SG1-64beb5e4-c0"}} |
| repeat_actions | True |
| resource_type | instance |
| severity | low |
| state | insufficient data |
| state_reason | Not evaluated yet |
| state_timestamp | 2018-07-20T06:00:33.142762 |
| threshold | 10.0 |
| time_constraints | [] |
| timestamp | 2018-07-20T06:00:33.142762 |
| type | gnocchi_aggregation_by_resources_threshold |
| user_id | 61fb5c6193e549f3baee26bd508c0b29 |
+---------------------------+------------------------------------------------------------------------------------------------------------------+How to trigger alarms:
As shown in the above Ceilometer command, alarm state is shown as "insufficient data". Alarm is triggered by Ceilometer once alarm state changes to "alarm". To make VNF instance reach to the pre-defined threshold, some simple scripts could be used.
Note: Because Ceilometer pipeline set the default interval to 600s (10 mins), in order to reduce this interval, users could edit "interval" value in /etc/ceilometer/pipeline.yaml file and then restart Ceilometer service.
Another way could be used to check if backend action is handled well in Tacker:
curl -H "Content-Type: application/json" -X POST -d '{"alarm_id": "35a80852-e24f-46ed-bd34-e2f831d00172", "current": "alarm"}' http://ubuntu:9890/v1.0/vnfs/7582cdf4-58ed-4df8-8fa2-c15938adf70b/vdu_lcpu_usage_scaling_in/SP1-in/v2fq7rd7Then, users can check Horizon to know if vnf is respawned. Please note that the url used in the above command could be captured from "ceilometer alarm-show command as shown before. "key" attribute in body request need to be captured from the url. The reason is that key will be authenticated so that the url is requested only one time.