Skip to content

cc4i/llm-alerting

Repository files navigation

An intelligent alerting with Gemini & Function calling

Description

Leveraging LLMs on GCP to gernerate meaningful guidance for each alerting in order to accelerate problem solving rather than a simple alerting message.

Architecture

architecture

Prerequisite

Deployment

# 1. Modify initial setting in bin/config.env
#
# Notice: 
# To replace webhook url, I use sed here. In sed's replacement section, the '&' represents the entire text that matched your search pattern. To include a literal '&' in the replacement, you need to escape it with a backslash.
#

vi bin/config.env

# 2. Update all related parametes.
bin/config.sh var

# 3. Deploy llm-alerting Cloud Run.
bin/config.sh run

# 4. Create a GKE cluster for demo (Ignore this step if using existed GKE)
bin/config.sh demo

# 5. Create a notification channel and alerting policy for demo.
config.sh alert

Advanced

Make sure your have enough quota for Vertex AI to run more Funtion Calling, such as main-ext.py. Otherwise you would meet following error:


google.api_core.exceptions.ResourceExhausted: 429 Quota exceeded for aiplatform.googleapis.com/generate_content_requests_per_minute_per_project_per_base_model with base model: gemini-pro. Please submit a quota increase request. https://cloud.google.com/vertex-ai/docs/generative-ai/quotas-genai.