Skip to content

PerformanceTuning.md

Jump to bottom
Chris edited this page Jun 22, 2025 · 3 revisions

Performance & Scalability

DockFlare includes several features designed to ensure it runs efficiently, provides a responsive user experience, and respects Cloudflare API rate limits, especially in environments with many containers.

1. API Rate Limit Protection

These features are designed to minimize unnecessary API calls and prevent errors from Cloudflare.

Concurrent Operation Limiting

  • Problem: When many containers start or stop at once, DockFlare could try to create or delete numerous DNS records simultaneously, potentially hitting Cloudflare's API rate limits.
  • Solution: DockFlare uses a semaphore to limit the number of DNS operations that can run in parallel.
  • Configuration:
    • MAX_CONCURRENT_DNS_OPS: The environment variable controlling the maximum number of simultaneous DNS operations.
    • Default: 3. You can cautiously increase this if you have high API limits or decrease it to 1 or 2 if you encounter rate-limiting errors.

Intelligent Configuration Updates

  • Problem: Pushing the entire tunnel configuration to Cloudflare on every minor change is inefficient.
  • Solution: Before sending an update, DockFlare generates a "fingerprint" of its desired ingress rule configuration and compares it to the fingerprint of the configuration currently live on Cloudflare. An API PUT request is only sent if a change is actually detected.
  • Benefit: This drastically reduces API traffic and speeds up operations, as no updates are sent if the state is already consistent.

API Call Caching

  • Problem: Repeatedly asking for the same information, like a Zone ID for a domain, is redundant.
  • Solution: DockFlare caches frequently requested, non-volatile information (like Zone IDs) in memory for a short period.
  • Benefit: This reduces latency and the number of GET requests made to the Cloudflare API.

2. Efficient Startup & Reconciliation

These features ensure the application starts quickly and handles large numbers of containers without getting bogged down.

Batched Initial Scan

  • Problem: On startup, scanning hundreds of containers sequentially can be slow.
  • Solution: DockFlare processes the initial list of containers in small batches. This allows the application to remain responsive and provide progress feedback in the UI during a large initial scan.
  • Configuration:
    • RECONCILIATION_BATCH_SIZE: The environment variable controlling how many containers are processed in each batch during the startup scan.
    • Default: 5.

Asynchronous Initialization

  • Problem: Waiting for all initialization tasks (connecting to APIs, scanning containers, reconciling state) to complete before the UI is available can be inconvenient.
  • Solution: DockFlare's Web UI starts almost immediately, while the more intensive initialization and reconciliation tasks run in a background thread.
  • Benefit: You can access the UI and see live logs right away, with status indicators showing the current state of the application (e.g., "Initializing", "Initial container scan...", "Running").

Clone this wiki locally