Skip to main content

SDK for communicating with Mona's servers

Project description

Mona Python SDK

Mona's logo

Mona’s SDK is a python based package which enables you to securely access Mona’s API and export your data to Mona directly from within your code, either message-by-message or as a batch.

Installation

$ pip install mona_sdk

Quick Start and Example

  1. Install as described above.
  2. Get an "admin"-roled API key and secret from your dashboard.
  3. Set environment variables as mentioned below.
  4. Instrument code with Mona's client as seen below.

Data exporting arguments:

We recommend you get acquainted with Mona's key concepts in our docs.

MonaSingleMessage: A dataclass wrapping all needed fields for a Mona data exporting message, containing the following:

  • contextClass (str): (Required) The name of the context class to which you are currently exporting data.
  • message (dict): (Required) A JSON-serializable dict containing all relevant monitoring information to send to Mona's servers.
  • contextId (str): (Optional) A unique identifier for the current context instance. One can export multiple messages with the same context id and Mona would aggregate all of these messages to one big message on its backend. If none is given, Mona will create a random uuid for it. This is highly unrecommended - since it takes away the option to update this data in the future.
  • exportTimestamp (int | str): (Optional) This is the primary timestamp Mona will use when considering the data being sent. It should be a date (ISO string, or a Unix time number) representing the time the message was created. If this field isn't provided, the message exportTimestamp will be the time in which the exporting function was called.
  • action (str): (Optional) The action Mona should do with the message to an existing context:
    • "OVERWRITE": (default) The values in the given fields will replace values already existing in the given fields.
    • "ADD": The values in the given fields will be added to the values already existing in these fields (will be eventually treated as arrays of values).
    • "NEW": completely reset the entire record of the given context id to only refer to the given message, so everything before this message is no longer relevant.
from mona_sdk.client import Client, MonaSingleMessage
import time

api_key = <An API key is accessible in the admin page in your dashboard>
secret = <secret corresponding to the given api_key>

my_mona_client = Client(api_key, secret)

# One can send a single message to Mona's servers by calling export() with a 
# MonaSingleMessage object:
succeed_to_export = my_mona_client.export(MonaSingleMessage(
    message={'monitoring_information_1': '1', 'monitoring_information_2': '2'}, 
    contextClass='MY_CONTEXT_CLASS_NAME', 
    contextId='CONTEXT_INSTANCE_UNIQUE_ID', 
    exportTimestamp=time.time(),
    action="OVERWRITE"
))

# Another option is to send a batch of messages to Mona using export_batch:
messages_batch_to_mona = []
for context_instance in my_data:
    messages_batch_to_mona.append(
            MonaSingleMessage(
                message=context_instance.relevant_monitoring_information,
                contextClass="MY_CONTEXT_CLASS_NAME",
                contextId=context_instance.unique_id,
                exportTimestamp=context_instance.timestamp,
            )
        )
        
# Use dafault_action to select a default action for messages with no specified action.
export_result = my_mona_client.export_batch(
    messages_batch_to_mona, 
    default_action="ADD",
    )

Mona SDK services

Mona sdk provides a simple API to access your information and control your configuration and data on Mona. You can see all functions info and examples on our docs under REST API.

The available services are:

get_suggested_config

validate_config

upload_config_per_context_class

get_config_history

get_insights

get_ingested_data_for_a_specific_segment

get_suggested_config_from_user_input

get_aggregated_data_of_a_specific_segment

get_aggregated_stats_of_a_specific_segmentation

Get your current Mona configuration:

my_current_mona_config = my_client.get_config()

Upload a new configuration:

Arguments:

  • config: Your new Mona configuration represented as a python dict (both the configuration dict with your user_id as the top key and just the configuration dict itself are accepted).
  • commit_message: A short description of the changes that were made.
  • author: An email address identifying the configuration uploader. Mona will use this mail to send updates regarding re-creation of insights upon this configuration change. When not supplied, the author will be the Client's api-key, and you will not get updates regarding the changes mentioned above. Must be provided when using un-authenticated mode.
new_configuration = {
    "MY_CONTEXT_CLASS": {
        "fields": <fields dict>, 
        "field_vectors": <field vectors dict>
        "stanzas_global_defaults": <global defaults dict>, 
        "stanzas": <stansas dict>,
        "notifications": <notifications dict>
    }
}
author = 'author@mycompany.io'
upload_result = my_client.upload_config(new_configuration, "My commit message", author)

# the return value format will be:
# upload_result == {
#    "success": <was the upload successful>, (bool)
#    "new_config_id": <the new configuration ID> (str)
#}

Environment variables

Mona uses several environment variables you can set as you prefer:

  • MONA_SDK_RAISE_AUTHENTICATION_EXCEPTIONS - Set to true if you would like Mona's client to raise authentication related exceptions. When set to false and such an exception is met, every function call will return false. Use client.is_active() in order to check authentication status. (default value: False).
  • MONA_SDK_RAISE_EXPORT_EXCEPTIONS - set to true if you would like Mona's client to raise export related exceptions. When set to false and an export (or part of it) fails, the failure reason will be logged (default value: False).
  • MONA_SDK_RAISE_SERVICE_EXCEPTIONS - set to true if you would like Mona's client to raise services requests related exceptions (see all available services under Mona SDK services section in this doc). When set to false, and such an exception is met, the function will log an error and return false (default value: False).
  • MONA_SDK_NUM_OF_RETRIES_FOR_AUTHENTICATION - Number of retries to authenticate in case Mona's client unexpectedly cannot get an authentication response from the server (default value: 3).
  • MONA_SDK_WAIT_TIME_FOR_AUTHENTICATION_RETRIES_SEC - Number of seconds to wait between every authentication retry (default value: 2).
  • MONA_SDK_SHOULD_LOG_FAILED_MESSAGES - When true, failed messages will be logged ("ERROR" level).
  • MONA_SDK_OVERRIDE_APP_SERVER_HOST - When provided, all configuration related calls to mona's servers will use this host name instead of the default one ("api<user_id>.monalabs.io").
  • MONA_SDK_OVERRIDE_REST_API_HOST- When provided, all messages (data export) to mona's rest-api will use this host address instead of the default one ("incoming<user_id>.monalabs.io").
  • MONA_SDK_OVERRIDE_REST_API_URL- When provided, all messages to mona's rest-api will use this full url address (including "http" prefix and endpoints suffix) instead of the default one. Note: this is mostly for internal use, please use MONA_SDK_OVERRIDE_REST_API_HOST if needed.
  • MONA_SDK_SHOULD_USE_SSL - Should the communication with Mona's servers be with (https) or without (http) ssl (default value: True).
  • MONA_SDK_SHOULD_USE_AUTHENTICATION - When set to false, the communication with Mona's servers will not use authentication (user_id should be provided to the Client constructor instead of an api_key and a secret), Note: this mode is not supported on the servers by default, and must be explicitly requested from Mona's team (default value: True).
  • MONA_SDK_FILTER_NONE_FIELDS_ON_EXPORT - When set to true, Mona's client will filter out all fields with None values from the message dict to export (default value: False). Note that passing a None value may be required in order to delete a pre-existing value. To allow None values, use filter_none_fields=False which overrides this parameter both in export() and export_batch() functions.
  • MONA_SDK_DEFAULT_SAMPLING_FACTOR - A float in the range [0, 1], which sets the random client-side sampling done by the SDK before sending the data into Mona servers. If this value is less than 1, only a random (see below) sample of the given proportion is actually going to be sent, leaving the rest of the data unattended. Use with caution. (random - using hashing with sha224 on the context id, if supplied, or by random.random() otherwise.)
  • MONA_SDK_SAMPLING_CONFIG - Allows to override the sampling factor (see MONA_SDK_DEFAULT_SAMPLING_FACTOR above) by context class. If set, the expected format is a valid JSON-object string. Keys are the names of the context classes to override, and the value is expected to be floats in the range of [0, 1]. For example: '{"class1": 0.3, "class2": 0.5, "class3": 1}'

Another way to control these behaviors is to pass the relevant arguments to the client constructor as follows (the environment variables are used as defaults for these arguments, and by passing these arguments to the constructor you can override the default values you set with the environment variables):

my_mona_client = Client(
    api_key,
    secret,
    raise_authentication_exceptions=True,
    raise_export_exceptions=True,
    raise_service_exceptions=True,
    num_of_retries_for_authentication=6,
    wait_time_for_authentication_retries=0,
    should_log_failed_messages=True,
    filter_none_fields_on_export=True,
    default_sampling_rate=0.1,
    context_class_to_sampling_rate={"class1": 0.5, "class2": 1},
)

Logging

Unrelated to the actual data being exported, Mona's client may log debug/info/warning/error for various reasons, including to help with debugging communications with Mona's server. To make logging as adaptable to your system as possible, Mona is using its own logger named "mona-logger". You can configure it in your code by just calling

logging.getLogger("mona-logger")

and then setting handlers and formatters as you please.

You can also configure Mona's logging using the following environment variables:

  1. MONA_SDK_LOGGING_LEVEL - set this to the wanted level, according to python's logging constants:

    • "CRITICAL" (50)
    • "ERROR" (40)
    • "WARNING" (30)
    • "INFO" (20)
    • "DEBUG" (10)
    • "NOTSET" (0)
  2. MONA_SDK_LOGGER_NAME - you can change Mona's logger name.

Special field names

Avoid using field names with "MONA" as their prefix, Mona uses this pattern internally. If you do that, these fields will be changed to a have prefix of "MY_MONA" before exporting to Mona's servers.

Testing the client code

The client's tests are written using unittest framework.

In order to run the tests type the following to your shell:

python -m unittest mona_sdk/tests/client_tests.py

Asynchronous Client

Same services as the regular client are provided in a non-blocking version in the AsyncClient, in order to give you the option to embed the code you write using this package with asyncIO code.

AsyncClient Constructor parameters:

  • event_loop (_UnixSelctorEventLoop): (optional) The event loop that will manage the threads. If not provided, a default is used.
  • executor (TreadPoolExecutor): (optional) The executor the will manage the thread pool. If not provided, a default is used.

When using AsyncClient, while all the regular (synchronous) client functions are still supported, you can simply add "_async" suffix to any function (e.g export_async() instead of export(); export_batch_async() instead of export_batch() etc). The async version of the methods accept the same parameters as the synchronous version, in addition to the following parameters:

  • event_loop (_UnixSelctorEventLoop): (optional) This overrides the event loop provided for the AsyncClient constructor.
  • executor (TreadPoolExecutor): (optional) This overrides the executor provided for the AsyncClient constructor.

An example for using export_batch_async to send data to Mona asynchronously:

from mona_sdk.client import AsyncClient, MonaSingleMessage
import asyncio

def main():
    api_key = <An API key is accessible in the admin page in your dashboard>
    secret = <secret corresponding to the given api_key>

    my_mona_async_client = AsyncClient(api_key, secret)

    messages_batch_to_mona = []
    for context_instance in my_data:
        messages_batch_to_mona.append(
            MonaSingleMessage(
                message=context_instance.relevant_monitoring_information,
                contextClass="MY_CONTEXT_CLASS_NAME",
                contextId=context_instance.unique_id,
                exportTimestamp=context_instance.timestamp,
            )
        )

    asyncio.create_task(my_mona_async_client.export_batch_async(messages_batch_to_mona))

asyncio.run(main())

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mona_sdk-0.0.39.tar.gz (30.7 kB view hashes)

Uploaded source

Built Distribution

mona_sdk-0.0.39-py3-none-any.whl (28.9 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page