Feat dockerfile

[ArnabChatterjee20k]

What does this PR do?

(Provide a description of what this PR does.)

Test Plan

(Write your test plan here. If you changed any code, please provide us with clear instructions on how you verified your changes work.)

Related PRs and Issues

(If this PR is related to any other PR or resolves any issue or related to any issue link all related PR and issues here.)

Have you read the Contributing Guidelines on issues?

(Write your answer here.)

Summary by CodeRabbit

• New Features
  • Added embedding model support with optimized caching for improved performance.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Warning

Rate limit exceeded

@ArnabChatterjee20k has exceeded the limit for the number of commits or files that can be reviewed per hour. Please wait 21 minutes and 59 seconds before requesting another review.

⌛ How to resolve this issue?

After the wait time has elapsed, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout.

Please see our FAQ for further information.

📥 Commits

Reviewing files that changed from the base of the PR and between 47e9b53 and 871c29e.

📒 Files selected for processing (1)

• docker-compose.yml (1 hunks)

Walkthrough

This change introduces a new embedding/LLM service infrastructure using Ollama. The .env file specifies the embedding models to be used. The docker-compose.yml file defines a new ollama service with build arguments, port exposure, and persistent volume storage for cached models. The Dockerfile configures the ollama image to preload specified models during build, cache them in Docker layers, and verify their availability at container startup, with a background check loop to pull missing models if needed.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

• Dockerfile background process logic: The entry point script contains a loop that sleeps and checks model availability, with conditional pulling of missing models. The build phase runs Ollama in the background with error handling to ensure models are cached correctly.
• Cross-file integration: Understanding how the .env variable flows through docker-compose.yml to the Dockerfile ARG requires tracing configuration through all three files.
• Model persistence and caching: The volume mounting strategy (appwrite-models at /root/.ollama) needs verification to ensure models persist across restarts and the caching behavior works as intended.
• Startup reliability: The verification loop at container startup and the error handling during model pulling should be reviewed for potential race conditions or edge cases where models fail to pull.

Pre-merge checks and finishing touches

❌ Failed checks (1 inconclusive)

Check name	Status	Explanation	Resolution
Title check	❓ Inconclusive	The title 'Feat dockerfile' is vague and generic. While it references 'dockerfile', it doesn't clearly convey what the actual change accomplishes or why it matters.	Use a more descriptive title that explains the purpose, such as 'Add Ollama service with model pre-caching' or 'Set up Docker Compose with Ollama and embedding models'.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 3

🧹 Nitpick comments (3)

.env (1)

1-1: Add a blank line at the end of the file.

This file is missing a trailing newline, which is flagged by dotenv-linter and is a POSIX standard best practice.

 _APP_EMBEDDING_MODELS=embeddinggemma
+

docker-compose.yml (2)

4-18: Add a health check to verify Ollama is responsive.

Without a health check, the container may report as running even if the Ollama service is unresponsive. This is critical for ensuring reliable service availability.

   services:
     ollama:
       build:
         context: .
         args:
           MODELS: ${_APP_EMBEDDING_MODELS:-embeddinggemma}
           OLLAMA_KEEP_ALIVE: 24h
       restart: unless-stopped
+      healthcheck:
+        test: ["CMD", "curl", "-f", "http://localhost:11434/api/tags"]
+        interval: 30s
+        timeout: 5s
+        retries: 3
+        start_period: 10s
       ports:

---

1-18: Document the interaction between build-time model preloading and persistent volume caching.

The Dockerfile preloads models during build (cached in layers), and docker-compose mounts a persistent volume for /root/.ollama. If MODELS changes and the image is rebuilt, the old cached models in the volume won't be automatically cleaned up. Consider documenting this behavior and/or adding a note about when to rebuild vs. when to recreate volumes.

📜 Review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between d3cb8f1 and 47e9b53.

📒 Files selected for processing (3)

• .env (1 hunks)
• docker-compose.yml (1 hunks)
• dockerfile (1 hunks)

🧰 Additional context used

🪛 dotenv-linter (4.0.0)

.env

[warning] 1-1: [EndingBlankLine] No blank line at the end of the file

(EndingBlankLine)

[coderabbitai]

🛠️ Refactor suggestion | 🟠 Major

Quote the $MODELS variable in the build-time model loop.

Unquoted variables risk word-splitting if model names contain spaces or special characters. Also use word-boundary matching in grep.

   for m in $MODELS; do \
       echo "Pulling model $m..."; \
-      ollama pull $m || exit 1; \
+      ollama pull "$m" || exit 1; \

📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change

	for m in $MODELS; do \
	echo "Pulling model $m..."; \
	ollama pull $m || exit 1; \
	done && \
	pkill ollama
	for m in $MODELS; do \
	echo "Pulling model $m..."; \
	ollama pull "$m" || exit 1; \
	done && \
	pkill ollama

[coderabbitai]

⚠️ Potential issue | 🔴 Critical

Fix the ENTRYPOINT command logic—$0 won't pass CMD correctly.

In the bash -c context, $0 refers to bash itself, not the CMD argument. This means exec ollama $0 passes "bash" to ollama (invalid), and CMD ["serve"] is never invoked. The background model check also won't block container startup if a pull fails, so the service may start with missing models.

-ENTRYPOINT ["/bin/bash", "-c", "(sleep 2; for m in $MODELS; do ollama list | grep -q $m || ollama pull $m; done) & exec ollama $0"]
+ENTRYPOINT ["/bin/bash", "-c", "for m in \"$MODELS\"; do ollama list | grep -qw \"$m\" || ollama pull \"$m\" || exit 1; done && exec ollama serve"]
 CMD ["serve"]

This change:

• Removes the $0 reference and explicitly uses serve.
• Makes the model check synchronous so the container fails if models are unavailable.
• Adds word-boundary matching to grep (-w) and quotes variables to handle model names safely.
• Removes the unused CMD since the command is now fully specified in ENTRYPOINT.

Committable suggestion skipped: line range outside the PR's diff.

🤖 Prompt for AI Agents

In dockerfile around line 22, the ENTRYPOINT uses bash -c with $0 (which is
bash) and runs the model-check in background so CMD isn't passed and failed
pulls are ignored; change ENTRYPOINT to run a synchronous shell script/command
that quotes $MODELS, iterates over each model, uses word-boundary grep (-w)
and/or exact matching, attempts ollama pull and exits non-zero on any pull
failure so startup fails if models are missing, then exec ollama serve
explicitly (remove reliance on $0 and remove the now-unused CMD).