IM probe

1 INTRODUCTION

Infrastructure Manager (IM) is a tool that eases access and usability of IaaS clouds by automating virtual machine infrastructure selection, deployment, configuration, software installation, monitoring and update of virtual appliances. In order to offer relevant data to users about the health of its operations, the most relevant metrics are obtained by a python based monitor agent, using the IM REST API. Once obtained data from platform, this content is processed and sent to a Zabbix server which store and visualize monitorization data.

2 ARCHITECTURE

The monitorization agent is a client-side software, that connects with IM to capture monitorization data, and sends it to Zabbix server. Agent does not need any open ports, just send data on passive mode to Zabbix and gather data from IM via IM API REST by GET, POST, PUT and DELETE http requests. Agent is composed by scripts that use IM API and Zabbix API, and loads configurations from files that must be edited correctly by an administrator.

3 FILE STRUCTURE

The agent is composed by a principal python script named probeim.py, four support libraries named loadconfigs.py, tokenmng.py, IMinfrastructureOper.py and zabbix.py and a directory conf with two configuration files, imzabbix.conf and authorizationHeader.txt.

├── conf
│   ├── imzabbix.conf
│   └── authorizationHeader.txt
├── loadconfigs.py
├── probeim.py
├── IMinfrastructureOper.py
├── tokenmng.py
└── zabbix.py

3.1 SCRIPTS

• File probeim.py is the executable script that gathers all measured data and sends it to zabbix.
• File zabbix.py contains functionalities related to interaction with zabbix API. Here we can authenticate against zabbix obtaining a valid token; get a list of all monitored items; create items; get template and item identifiers.
• File IMinfrastructureOper.py contains functionalities related to interaction with IM environment through its API. Here we can throw actions like creating infrastructure, listing infrastructure, creating VM, deleting infrastructure, and get the status of each of these actions.
• File tokenmng.py has functions that connect with IAM to request a token from a refresh token, so client remains authenticated to be able to launch operations over IM.
• File loadconfigs.py contains classes that help to bring configuration data to the agent and are used to connect to Zabbix, IAM and IM.

3.2 CONFIGURATION

• imzabbix.conf

Example from /conf here

Section	zabbix
uri	the api url to make requests
username	the zabbix username used for authentication in order to get a token
password	the zabbix password used for authentication
server_add	IP address where zabbix server is running
agent_delay	Time in seconds for each loop that catches data
template	the name of template used for IM measurements
monitoredhost	the hostname of the monitored host

Section	im
urlbase	the IM api url to make requests
radl	the radl configuration for IM to create a VM.

Section	iam
urlrefresh	the api url to request a new token from a refresh token

Section	log
loglevel	the severity of logs. This can take the following values: ERROR, WARNING, INFO, DEBUG

Default values is WARNING. |

• authorizationHeader.txt

Example from /conf here

4 DATA SOURCES

METHOD /URL	Items
GET /infrastructures	List_inf
POST /infrastructures	Create_inf
PUT /infrastructures//start	Start_inf
POST /infrastructures/

body: RADL document | Create_vm | | DELETE /infrastructures/<infId|Delete_inf |

To connect to IM an appropriated header must be set.

HEADERS = {
        "Content-Type" : "text/plain",
        "Accept": "application/json",
        "Authorization" : authoriz
    }

Where the authoriz field comes from authorizationHeader.txt file.

5 DOCKER CONTAINER

Create or download a Dockerfile Use your Docker user and a name you want for your image.

Download and execute the script used to request to IM an access token. You need your client credentials, IAM user and password.

[root@localhost imzabbix]# sh get-access-token.sh 

{"access_token":"eyJraWQiOiJyc2E...","token_type":"Bearer","refresh_token":"eyJhbGciOiJub25lIn0.eyJqd...","expires_in":3599,"scope":"address phone openid email profile offline_access","id_token":"eyJraWQiOi..."}
[root@localhost imzabbix]#

Build an image from the directory where your Dockerfile is located.

[root@localhost imzabbix]# ls Dockerfile 
Dockerfile
[root@localhost imzabbix]# docker build -t <dockeruser>/<my_built_image> .
Sending build context to Docker daemon 692.2 kB
Step 1 : FROM centos
 ---> 0584b3d2cf6d
...
Successfully built afe963948e10
[root@localhost imzabbix]#

After image was created, set up required parameters and run the docker container.

[root@localhost imzabbix]# ZABBIX_PASSWORD=zabbix
[root@localhost imzabbix]# ZABBIX_SERVER=172.17.0.3
[root@localhost imzabbix]# ZABBIX_API_URL=http://172.17.0.4/api_jsonrpc.php
[root@localhost imzabbix]# ZABBIX_USER=Admin
[root@localhost imzabbix]# CLIENT_ID=d573b18e-7a59-4bed-bbbb-ebad2b5f8299
[root@localhost imzabbix]# CLIENT_SECRET=NxFfYvk19hiKuuLo50OO44nuwShRsV45UrvivWUwvF-4Szdadu2fuh7BwHLuttbZFeQYhYmEbbxdfSDrJajdNg
[root@localhost imzabbix]# TOKEN=eyJraWQiOiJyc2E...
[root@localhost imzabbix]# REFRESH=REFRESH=eyJhbGciOiJub25lIn0.eyJqd...
[root@localhost imzabbix]#
[root@localhost imzabbix]# docker run --name MY_IMZABBIX_CONTAINER -e CLIENT_ID=$CLIENT_ID -e CLIENT_SECRET=$CLIENT_SECRET -e TOKEN=$TOKEN -e REFRESH=$REFRESH -e ZABBIX_USER=$ZABBIX_USER -e ZABBIX_PASSWORD=$ZABBIX_PASSWORD -e ZABBIX_SERVER=$ZABBIX_SERVER -e ZABBIX_API_URL=$ZABBIX_API_URL -d <dockeruser>/<my_built_image>