DOCSEARCHbyAlgolia
Submit Docs

Oodle Documentation

website·Indexed 34 days ago

https://docs.oodle.ai

A platform for LLM application evaluation and agent observability. Oodle provides tools for creating datasets, running experiments with evaluators, testing agents via webhooks, and tracking costs and traces using OpenTelemetry Gen AI semantic conventions. It includes a Playground for prompt testing, a CLI, a Terraform provider, and an MCP server for managing webhooks.

Tokens
186.5K
Snippets
578
Records
1.6K
Agent score
66%

What's inside Oodle

  1. Overview of Oodle MCP server observability tools

    The Oodle MCP server provides a suite of observability tools categorized by the following capabilities:

    • Logs: Query, filter, and aggregate logs using flexible time ranges and field-based filtering.
    • Metrics: Execute PromQL queries, search for metrics, and explore labels.
    • Traces: Search distributed traces by service, operation, or custom tags and retrieve trace attributes.
    • Alerts: Access active alerts, view monitor configurations, and analyze historical trigger data.
    • Grafana: Programmatically search, retrieve, and update dashboards or query panel data. Note: Saving or updating dashboards requires an Editor or Admin API key.
    • RUM (Real User Monitoring): Query user sessions, browse session events, and investigate frontend issues reported by the RUM SDK.
    • Integrations: Discover available integrations and retrieve setup specifications for onboarding new data sources.

  2. Overview of Oodle Synthetic Monitoring

    Synthetic monitoring in Oodle allows for proactive monitoring of service availability and performance by running automated checks from multiple geographic locations. Unlike infrastructure-based monitoring, synthetic monitoring simulates real user interactions to test services from an external perspective.

  3. Overview of Oodle Kubernetes Monitoring

    Oodle includes built-in Kubernetes monitoring capabilities that provide visibility into cluster health, resource utilization, pod status, and workload performance. A key advantage of this integrated monitoring is that it allows users to track these metrics without needing to write PromQL queries.

  4. Overview of Oodle Incident Bot

    Oodle Incident Bot is an AI-powered incident response assistant integrated with Slack incident channels. It maintains shared context across a team, allowing responders to ask questions that build upon previous discoveries, eliminating the need to repeatedly provide context to the AI during an incident.

  5. Overview of Oodle Database Monitoring

    Oodle Database Monitoring provides visibility into database performance through a single interface. It allows developers and operators to track query execution times, identify slow queries, monitor host resource utilization, and analyze query samples.

  6. Overview of Oodle Observability Platform

    Oodle is a fully managed, serverless observability platform designed as a drop-in replacement for Prometheus and the ELK stack at scale. It is built on top of S3 to provide enterprise-grade observability at a lower cost than traditional enterprise offerings. Oodle focuses on AI-powered 'dashboard-less debugging' to accelerate incident resolution and maintains full compatibility with OpenTelemetry, Prometheus, Grafana, and the ELK stack.

  7. Overview of Oodle Logs Visualization

    Oodle's logs system utilizes a custom columnar storage engine designed to handle high-volume queries without timeouts. The visualization interface provides AI-powered analysis, automatic pattern detection, and direct integration with metrics and traces.

  8. Overview of Oodle MCP Server Observability Tools

    The Oodle MCP server provides a suite of observability tools categorized by the following capabilities:

    • Logs: Query and filter logs using flexible time ranges and field-based filtering, discover available fields, and run aggregations.
    • Metrics: Execute PromQL queries to access metrics data, search for metrics, explore labels, and validate queries.
    • Traces: Search distributed traces by service, operation, or custom tags, and retrieve specific trace details and attributes.
    • Alerts: Access active alerts, view monitor configurations, and analyze historical trigger data.
    • Grafana: Programmatically search, retrieve, and update dashboards, and query panel data. Note: Dashboard write operations (save/update) require an Editor or Admin API key.
    • RUM (Real User Monitoring): Query user sessions, browse session events, and investigate frontend issues/errors reported by the RUM SDK.
    • Integrations: Discover available integrations and retrieve setup specifications for onboarding new data sources.

  9. Overview of Oodle MCP Server Observability Tools

    The Oodle MCP server provides a suite of observability tools categorized by the following capabilities:

    • Logs: Query and filter logs using flexible time ranges and field-based filtering, discover available fields, and run aggregations.
    • Metrics: Execute PromQL queries to access metrics data, search for metrics, and explore labels.
    • Traces: Search distributed traces by service, operation, or custom tags, and retrieve specific trace attributes.
    • Alerts: Access active alerts, view monitor configurations, and analyze historical trigger data.
    • Grafana: Programmatically search, retrieve, and update dashboards or query panel data. Note: Saving or updating dashboards requires an Editor or Admin API key.
    • RUM (Real User Monitoring): Query user sessions, browse session events, and investigate frontend issues reported by the RUM SDK.
    • Integrations: Discover available integrations and retrieve setup specifications for onboarding new data sources.

  10. Overview of Oodle MCP server observability tools

    The Oodle MCP server provides a suite of observability tools categorized by the following capabilities:

    • Logs: Query and filter logs using flexible time ranges and field-based filtering, discover fields, and run aggregations.
    • Metrics: Execute PromQL queries, search for metrics, explore labels, and validate queries.
    • Traces: Search distributed traces by service, operation, or custom tags, and retrieve specific trace details and attributes.
    • Alerts: Access active alerts, view monitor configurations, and analyze historical trigger data.
    • Grafana: Programmatically search, retrieve, and update dashboards or query panel data. Note: Dashboard write operations (save/update) require an Editor or Admin API key.
    • RUM (Real User Monitoring): Query user sessions, browse session events, and investigate frontend issues/errors reported by the RUM SDK.
    • Integrations: Discover available integrations and retrieve setup specifications for onboarding new data sources.

  11. Overview of Oodle Real User Monitoring (RUM)

    Oodle Real User Monitoring (RUM) provides end-to-end visibility into the actual user experience of a web application. It tracks key performance and stability metrics, including page load performance, JavaScript errors, user interactions, and provides full session replay capabilities.

  12. Overview of Oodle MCP server observability tools

    The Oodle MCP server provides a suite of observability tools categorized by the type of telemetry or management task. Available categories include:

    • Logs: Query, filter, and aggregate log patterns using flexible time ranges and field-based filtering.
    • Metrics: Execute PromQL queries, search for metrics, and explore labels.
    • Traces: Search distributed traces by service, operation, or custom tags and retrieve specific trace attributes.
    • Alerts: View active alerts, monitor configurations, and historical trigger data.
    • Grafana: Programmatically search, retrieve, and update dashboards or query panel data. Note: Writing/updating dashboards requires an Editor or Admin API key.
    • RUM (Real User Monitoring): Query user sessions, browse events, and investigate frontend errors reported by the RUM SDK.
    • Integrations: Discover available integrations and retrieve setup specifications for onboarding new data sources.