litellm-mirror/docs/my-website/docs/proxy/alerting.md

import Image from '@theme/IdealImage'; import Tabs from '@theme/Tabs'; import TabItem from '@theme/TabItem';

Alerting / Webhooks

Get alerts for:

Category	Alert Type
LLM Performance	Hanging API calls, Slow API calls, Failed API calls, Model outage alerting
Budget & Spend	Budget tracking per key/user, Soft budget alerts, Weekly & Monthly spend reports per Team/Tag
System Health	Failed database read/writes
Daily Reports	Top 5 slowest LLM deployments, Top 5 LLM deployments with most failed requests, Weekly & Monthly spend per Team/Tag

Works across:

• Slack
• Discord
• Microsoft Teams

Quick Start

Set up a slack alert channel to receive alerts from proxy.

Step 1: Add a Slack Webhook URL to env

Get a slack webhook url from https://api.slack.com/messaging/webhooks

You can also use Discord Webhooks, see here

Set SLACK_WEBHOOK_URL in your proxy env to enable Slack alerts.

export SLACK_WEBHOOK_URL="https://hooks.slack.com/services/<>/<>/<>"

Step 2: Setup Proxy

general_settings: 
    alerting: ["slack"]
    alerting_threshold: 300 # sends alerts if requests hang for 5min+ and responses take 5min+ 
    spend_report_frequency: "1d" # [Optional] set as 1d, 2d, 30d .... Specify how often you want a Spend Report to be sent
    
    # [OPTIONAL ALERTING ARGS]
    alerting_args:
        daily_report_frequency: 43200  # 12 hours in seconds
        report_check_interval: 3600    # 1 hour in seconds
        budget_alert_ttl: 86400        # 24 hours in seconds
        outage_alert_ttl: 60           # 1 minute in seconds
        region_outage_alert_ttl: 60    # 1 minute in seconds
        minor_outage_alert_threshold: 5 
        major_outage_alert_threshold: 10
        max_outage_alert_list_size: 1000
        log_to_console: false
    

Start proxy

$ litellm --config /path/to/config.yaml

Step 3: Test it!

curl -X GET 'http://0.0.0.0:4000/health/services?service=slack' \
-H 'Authorization: Bearer sk-1234'

Advanced

Redacting Messages from Alerts

By default alerts show the messages/input passed to the LLM. If you want to redact this from slack alerting set the following setting on your config

general_settings:
  alerting: ["slack"]
  alert_types: ["spend_reports"] 

litellm_settings:
  redact_messages_in_exceptions: True

Soft Budget Alerts for Virtual Keys

Use this to send an alert when a key/team is close to it's budget running out

Step 1. Create a virtual key with a soft budget

Set the soft_budget to 0.001

curl -X 'POST' \
  'http://localhost:4000/key/generate' \
  -H 'accept: application/json' \
  -H 'x-goog-api-key: sk-1234' \
  -H 'Content-Type: application/json' \
  -d '{
  "key_alias": "prod-app1",
  "team_id": "113c1a22-e347-4506-bfb2-b320230ea414",
  "soft_budget": 0.001
}'

Step 2. Send a request to the proxy with the virtual key

curl http://0.0.0.0:4000/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer sk-Nb5eCf427iewOlbxXIH4Ow" \
-d '{
  "model": "openai/gpt-4",
  "messages": [
    {
      "role": "user",
      "content": "this is a test request, write a short poem"
    }
  ]
}'

Step 3. Check slack for Expected Alert

<Image img={require('../../img/soft_budget_alert.png')}/>

Add Metadata to alerts

Add alerting metadata to proxy calls for debugging.

import openai
client = openai.OpenAI(
    api_key="anything",
    base_url="http://0.0.0.0:4000"
)

# request sent to model set on litellm proxy, `litellm --model`
response = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages = [], 
    extra_body={
        "metadata": {
            "alerting_metadata": {
                "hello": "world"
            }
        }
    }
)

Expected Response

<Image img={require('../../img/alerting_metadata.png')}/>

Select specific alert types

Set alert_types if you want to Opt into only specific alert types. When alert_types is not set, all Default Alert Types are enabled.

👉 See all alert types here

general_settings:
  alerting: ["slack"]
  alert_types: [
    "llm_exceptions",
    "llm_too_slow",
    "llm_requests_hanging",
    "budget_alerts",
    "spend_reports",
    "db_exceptions",
    "daily_reports",
    "cooldown_deployment",
    "new_model_added",
  ] 

Map slack channels to alert type

Use this if you want to set specific channels per alert type

This allows you to do the following

llm_exceptions -> go to slack channel #llm-exceptions
spend_reports -> go to slack channel #llm-spend-reports

Set alert_to_webhook_url on your config.yaml

model_list:
  - model_name: gpt-4
    litellm_params:
      model: openai/fake
      api_key: fake-key
      api_base: https://exampleopenaiendpoint-production.up.railway.app/

general_settings: 
  master_key: sk-1234
  alerting: ["slack"]
  alerting_threshold: 0.0001 # (Seconds) set an artificially low threshold for testing alerting
  alert_to_webhook_url: {
    "llm_exceptions": "https://hooks.slack.com/services/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",
    "llm_too_slow": "https://hooks.slack.com/services/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",
    "llm_requests_hanging": "https://hooks.slack.com/services/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",
    "budget_alerts": "https://hooks.slack.com/services/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",
    "db_exceptions": "https://hooks.slack.com/services/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",
    "daily_reports": "https://hooks.slack.com/services/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",
    "spend_reports": "https://hooks.slack.com/services/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",
    "cooldown_deployment": "https://hooks.slack.com/services/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",
    "new_model_added": "https://hooks.slack.com/services/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",
    "outage_alerts": "https://hooks.slack.com/services/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",
  }

litellm_settings:
  success_callback: ["langfuse"]

Provide multiple slack channels for a given alert type

model_list:
  - model_name: gpt-4
    litellm_params:
      model: openai/fake
      api_key: fake-key
      api_base: https://exampleopenaiendpoint-production.up.railway.app/

general_settings: 
  master_key: sk-1234
  alerting: ["slack"]
  alerting_threshold: 0.0001 # (Seconds) set an artificially low threshold for testing alerting
  alert_to_webhook_url: {
    "llm_exceptions": ["os.environ/SLACK_WEBHOOK_URL", "os.environ/SLACK_WEBHOOK_URL_2"],
    "llm_too_slow": ["https://webhook.site/7843a980-a494-4967-80fb-d502dbc16886", "https://webhook.site/28cfb179-f4fb-4408-8129-729ff55cf213"],
    "llm_requests_hanging": ["os.environ/SLACK_WEBHOOK_URL_5", "os.environ/SLACK_WEBHOOK_URL_6"],
    "budget_alerts": ["os.environ/SLACK_WEBHOOK_URL_7", "os.environ/SLACK_WEBHOOK_URL_8"],
    "db_exceptions": ["os.environ/SLACK_WEBHOOK_URL_9", "os.environ/SLACK_WEBHOOK_URL_10"],
    "daily_reports": ["os.environ/SLACK_WEBHOOK_URL_11", "os.environ/SLACK_WEBHOOK_URL_12"],
    "spend_reports": ["os.environ/SLACK_WEBHOOK_URL_13", "os.environ/SLACK_WEBHOOK_URL_14"],
    "cooldown_deployment": ["os.environ/SLACK_WEBHOOK_URL_15", "os.environ/SLACK_WEBHOOK_URL_16"],
    "new_model_added": ["os.environ/SLACK_WEBHOOK_URL_17", "os.environ/SLACK_WEBHOOK_URL_18"],
    "outage_alerts": ["os.environ/SLACK_WEBHOOK_URL_19", "os.environ/SLACK_WEBHOOK_URL_20"],
  }

litellm_settings:
  success_callback: ["langfuse"]

Test it - send a valid llm request - expect to see a llm_too_slow alert in it's own slack channel

curl -i http://localhost:4000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-1234" \
  -d '{
    "model": "gpt-4",
    "messages": [
      {"role": "user", "content": "Hello, Claude gm!"}
    ]
}'

MS Teams Webhooks

MS Teams provides a slack compatible webhook url that you can use for alerting

Quick Start

1. Get a webhook url for your Microsoft Teams channel

2. Add it to your .env

SLACK_WEBHOOK_URL="https://berriai.webhook.office.com/webhookb2/...6901/IncomingWebhook/b55fa0c2a48647be8e6effedcd540266/e04b1092-4a3e-44a2-ab6b-29a0a4854d1d"

3. Add it to your litellm config

model_list: 
    model_name: "azure-model"
    litellm_params:
        model: "azure/gpt-35-turbo"
        api_key: "my-bad-key" # 👈 bad key

general_settings: 
    alerting: ["slack"]
    alerting_threshold: 300 # sends alerts if requests hang for 5min+ and responses take 5min+ 

4. Run health check!

Call the proxy /health/services endpoint to test if your alerting connection is correctly setup.

curl --location 'http://0.0.0.0:4000/health/services?service=slack' \
--header 'Authorization: Bearer sk-1234'

Expected Response

<Image img={require('../../img/ms_teams_alerting.png')}/>

Discord Webhooks

Discord provides a slack compatible webhook url that you can use for alerting

Quick Start

1. Get a webhook url for your discord channel

2. Append /slack to your discord webhook - it should look like

"https://discord.com/api/webhooks/1240030362193760286/cTLWt5ATn1gKmcy_982rl5xmYHsrM1IWJdmCL1AyOmU9JdQXazrp8L1_PYgUtgxj8x4f/slack"

3. Add it to your litellm config

model_list: 
    model_name: "azure-model"
    litellm_params:
        model: "azure/gpt-35-turbo"
        api_key: "my-bad-key" # 👈 bad key

general_settings: 
    alerting: ["slack"]
    alerting_threshold: 300 # sends alerts if requests hang for 5min+ and responses take 5min+ 

environment_variables:
    SLACK_WEBHOOK_URL: "https://discord.com/api/webhooks/1240030362193760286/cTLWt5ATn1gKmcy_982rl5xmYHsrM1IWJdmCL1AyOmU9JdQXazrp8L1_PYgUtgxj8x4f/slack"

[BETA] Webhooks for Budget Alerts

Note: This is a beta feature, so the spec might change.

Set a webhook to get notified for budget alerts.

1. Setup config.yaml

Add url to your environment, for testing you can use a link from here

export WEBHOOK_URL="https://webhook.site/6ab090e8-c55f-4a23-b075-3209f5c57906"

Add 'webhook' to config.yaml

general_settings: 
  alerting: ["webhook"] # 👈 KEY CHANGE

2. Start proxy

litellm --config /path/to/config.yaml

# RUNNING on http://0.0.0.0:4000

3. Test it!

curl -X GET --location 'http://0.0.0.0:4000/health/services?service=webhook' \
--header 'Authorization: Bearer sk-1234'

Expected Response

{
  "spend": 1, # the spend for the 'event_group'
  "max_budget": 0, # the 'max_budget' set for the 'event_group'
  "token": "88dc28d0f030c55ed4ab77ed8faf098196cb1c05df778539800c9f1243fe6b4b",
  "user_id": "default_user_id",
  "team_id": null,
  "user_email": null,
  "key_alias": null,
  "projected_exceeded_data": null,
  "projected_spend": null,
  "event": "budget_crossed", # Literal["budget_crossed", "threshold_crossed", "projected_limit_exceeded"]
  "event_group": "user",
  "event_message": "User Budget: Budget Crossed"
}

API Spec for Webhook Event

• spend float: The current spend amount for the 'event_group'.

• max_budget float or null: The maximum allowed budget for the 'event_group'. null if not set.

• token str: A hashed value of the key, used for authentication or identification purposes.

• customer_id str or null: The ID of the customer associated with the event (optional).

• internal_user_id str or null: The ID of the internal user associated with the event (optional).

• team_id str or null: The ID of the team associated with the event (optional).

• user_email str or null: The email of the internal user associated with the event (optional).

• key_alias str or null: An alias for the key associated with the event (optional).

• projected_exceeded_date str or null: The date when the budget is projected to be exceeded, returned when 'soft_budget' is set for key (optional).

• projected_spend float or null: The projected spend amount, returned when 'soft_budget' is set for key (optional).

• event Literal["budget_crossed", "threshold_crossed", "projected_limit_exceeded"]: The type of event that triggered the webhook. Possible values are:

  • "spend_tracked": Emitted whenever spend is tracked for a customer id.
  • "budget_crossed": Indicates that the spend has exceeded the max budget.
  • "threshold_crossed": Indicates that spend has crossed a threshold (currently sent when 85% and 95% of budget is reached).
  • "projected_limit_exceeded": For "key" only - Indicates that the projected spend is expected to exceed the soft budget threshold.

• event_group Literal["customer", "internal_user", "key", "team", "proxy"]: The group associated with the event. Possible values are:

  • "customer": The event is related to a specific customer
  • "internal_user": The event is related to a specific internal user.
  • "key": The event is related to a specific key.
  • "team": The event is related to a team.
  • "proxy": The event is related to a proxy.

• event_message str: A human-readable description of the event.

Region-outage alerting (✨ Enterprise feature)

:::info Get a free 2-week license :::

Setup alerts if a provider region is having an outage.

general_settings:
    alerting: ["slack"]
    alert_types: ["region_outage_alerts"] 

By default this will trigger if multiple models in a region fail 5+ requests in 1 minute. '400' status code errors are not counted (i.e. BadRequestErrors).

Control thresholds with:

general_settings:
    alerting: ["slack"]
    alert_types: ["region_outage_alerts"] 
    alerting_args:
        region_outage_alert_ttl: 60 # time-window in seconds
        minor_outage_alert_threshold: 5 # number of errors to trigger a minor alert
        major_outage_alert_threshold: 10 # number of errors to trigger a major alert

All Possible Alert Types

👉 Here is how you can set specific alert types

LLM-related Alerts

Alert Type	Description	Default On
llm_exceptions	Alerts for LLM API exceptions	✅
llm_too_slow	Notifications for LLM responses slower than the set threshold	✅
llm_requests_hanging	Alerts for LLM requests that are not completing	✅
cooldown_deployment	Alerts when a deployment is put into cooldown	✅
new_model_added	Notifications when a new model is added to litellm proxy through /model/new	✅
outage_alerts	Alerts when a specific LLM deployment is facing an outage	✅
region_outage_alerts	Alerts when a specific LLM region is facing an outage. Example us-east-1	✅

Budget and Spend Alerts

Alert Type	Description	Default On
budget_alerts	Notifications related to budget limits or thresholds	✅
spend_reports	Periodic reports on spending across teams or tags	✅
failed_tracking_spend	Alerts when spend tracking fails	✅
daily_reports	Daily Spend reports	✅
fallback_reports	Weekly Reports on LLM fallback occurrences	✅

Database Alerts

Alert Type	Description	Default On
db_exceptions	Notifications for database-related exceptions	✅

Management Endpoint Alerts - Virtual Key, Team, Internal User

Alert Type	Description	Default On
new_virtual_key_created	Notifications when a new virtual key is created	❌
virtual_key_updated	Alerts when a virtual key is modified	❌
virtual_key_deleted	Notifications when a virtual key is removed	❌
new_team_created	Alerts for the creation of a new team	❌
team_updated	Notifications when team details are modified	❌
team_deleted	Alerts when a team is deleted	❌
new_internal_user_created	Notifications for new internal user accounts	❌
internal_user_updated	Alerts when an internal user's details are changed	❌
internal_user_deleted	Notifications when an internal user account is removed	❌

alerting_args Specification

Parameter	Default	Description
daily_report_frequency	43200 (12 hours)	Frequency of receiving deployment latency/failure reports in seconds
report_check_interval	3600 (1 hour)	How often to check if a report should be sent (background process) in seconds
budget_alert_ttl	86400 (24 hours)	Cache TTL for budget alerts to prevent spam when budget is crossed
outage_alert_ttl	60 (1 minute)	Time window for collecting model outage errors in seconds
region_outage_alert_ttl	60 (1 minute)	Time window for collecting region-based outage errors in seconds
minor_outage_alert_threshold	5	Number of errors that trigger a minor outage alert (400 errors not counted)
major_outage_alert_threshold	10	Number of errors that trigger a major outage alert (400 errors not counted)
max_outage_alert_list_size	1000	Maximum number of errors to store in cache per model/region
log_to_console	false	If true, prints alerting payload to console as a .warning log.