[PR #413] [MERGED] refact: Simple message queue refactoring #448

New Issue

2026-02-16T01:17:24-05:00

yindo commented

2026-02-16 01:17:24 -05:00

📋 Pull Request Information

Original PR: https://github.com/run-llama/llama_deploy/pull/413
Author: @masci
Created: 12/20/2024
Status: ✅ Merged
Merged: 12/20/2024
Merged by: @masci

Base: main ← Head: massi/simple-msg-queue

📝 Commits (10+)

6456be2 cosmetics
f33cf86 remove internal port concept
df82ffb derive from AbstractMessageQueue
41b0206 first draft
a49e2dc rewrite unit tests
483e614 fix tests
df1fb27 skip tool tests
680af0a fix e2e tests
b330aac fix linter
94d9de7 exclude untested code from coverage

📊 Changes

36 files changed (+591 additions, -962 deletions)

View changed files

📝 .pre-commit-config.yaml (+1 -1)
📝 e2e_tests/message_queues/message_queue_simple/test_message_queue.py (+2 -2)
📝 llama_deploy/__init__.py (+5 -2)
📝 llama_deploy/apiserver/deployment.py (+8 -7)
📝 llama_deploy/deploy/deploy.py (+4 -7)
📝 llama_deploy/message_consumers/base.py (+5 -3)
📝 llama_deploy/message_consumers/remote.py (+2 -1)
📝 llama_deploy/message_queues/__init__.py (+5 -3)
📝 llama_deploy/message_queues/base.py (+25 -42)
📝 llama_deploy/message_queues/simple/__init__.py (+4 -4)
📝 llama_deploy/message_queues/simple/client.py (+57 -85)
📝 llama_deploy/message_queues/simple/config.py (+12 -4)
📝 llama_deploy/message_queues/simple/server.py (+55 -269)
📝 llama_deploy/services/human.py (+8 -0)
📝 llama_deploy/services/tool.py (+12 -5)
📝 llama_deploy/services/workflow.py (+5 -13)
📝 pyproject.toml (+3 -2)
📝 tests/apiserver/test_deployment.py (+5 -5)
➕ tests/conftest.py (+17 -0)
📝 tests/control_plane/test_server.py (+7 -6)

...and 16 more files

📄 Description

Fixes #361

While working on #361 I found out it wasn't that easy to introduce the concept of topic in SimpleMessageQueue because there wasn't a clear separation between "client" and "server". With other flavors of message queues, the client is the one responsible for managing consumers, while the server is the actual broker service (Redis, Kafka and the likes). But the simple message queue also has a "server" part that acts as an in-memory storage.

With this PR, the server and the client are clearly separated, and only SimpleMessageQueue derives from AbstractMessageQueue, while the server class is just an object. This way, the simple message queue isn't special anymore, and it behaves exactly like the others.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/run-llama/llama_deploy/pull/413 **Author:** [@masci](https://github.com/masci) **Created:** 12/20/2024 **Status:** ✅ Merged **Merged:** 12/20/2024 **Merged by:** [@masci](https://github.com/masci) **Base:** `main` ← **Head:** `massi/simple-msg-queue` --- ### 📝 Commits (10+) - [`6456be2`](https://github.com/run-llama/llama_deploy/commit/6456be2b0a01d47c7dd45e183ee41ada2d2fcfc4) cosmetics - [`f33cf86`](https://github.com/run-llama/llama_deploy/commit/f33cf8624c8ab2e5b2613386d75089e920c431c0) remove internal port concept - [`df82ffb`](https://github.com/run-llama/llama_deploy/commit/df82ffbe35e8a9dc192f48ef006add37f23d76a1) derive from AbstractMessageQueue - [`41b0206`](https://github.com/run-llama/llama_deploy/commit/41b02062fa84b52634dba713f8abbe5ed23863aa) first draft - [`a49e2dc`](https://github.com/run-llama/llama_deploy/commit/a49e2dc6a8c1949ae0996302347215a342fd5909) rewrite unit tests - [`483e614`](https://github.com/run-llama/llama_deploy/commit/483e614e1f904ea170bd505eee3e523c3e5613ae) fix tests - [`df1fb27`](https://github.com/run-llama/llama_deploy/commit/df1fb27dd21954807d1d9acfb889d5686673de23) skip tool tests - [`680af0a`](https://github.com/run-llama/llama_deploy/commit/680af0a9fcfa6bd3b78769e25ac7f025d4c80bf4) fix e2e tests - [`b330aac`](https://github.com/run-llama/llama_deploy/commit/b330aacc26c4ed6f9a7f0a289a4e05123406ed72) fix linter - [`94d9de7`](https://github.com/run-llama/llama_deploy/commit/94d9de762d71c4cc3a310284a91fb0be3254414f) exclude untested code from coverage ### 📊 Changes **36 files changed** (+591 additions, -962 deletions) <details> <summary>View changed files</summary> 📝 `.pre-commit-config.yaml` (+1 -1) 📝 `e2e_tests/message_queues/message_queue_simple/test_message_queue.py` (+2 -2) 📝 `llama_deploy/__init__.py` (+5 -2) 📝 `llama_deploy/apiserver/deployment.py` (+8 -7) 📝 `llama_deploy/deploy/deploy.py` (+4 -7) 📝 `llama_deploy/message_consumers/base.py` (+5 -3) 📝 `llama_deploy/message_consumers/remote.py` (+2 -1) 📝 `llama_deploy/message_queues/__init__.py` (+5 -3) 📝 `llama_deploy/message_queues/base.py` (+25 -42) 📝 `llama_deploy/message_queues/simple/__init__.py` (+4 -4) 📝 `llama_deploy/message_queues/simple/client.py` (+57 -85) 📝 `llama_deploy/message_queues/simple/config.py` (+12 -4) 📝 `llama_deploy/message_queues/simple/server.py` (+55 -269) 📝 `llama_deploy/services/human.py` (+8 -0) 📝 `llama_deploy/services/tool.py` (+12 -5) 📝 `llama_deploy/services/workflow.py` (+5 -13) 📝 `pyproject.toml` (+3 -2) 📝 `tests/apiserver/test_deployment.py` (+5 -5) ➕ `tests/conftest.py` (+17 -0) 📝 `tests/control_plane/test_server.py` (+7 -6) _...and 16 more files_ </details> ### 📄 Description Fixes #361 While working on #361 I found out it wasn't that easy to introduce the concept of topic in `SimpleMessageQueue` because there wasn't a clear separation between "client" and "server". With other flavors of message queues, the client is the one responsible for managing consumers, while the server is the actual broker service (Redis, Kafka and the likes). But the simple message queue also has a "server" part that acts as an in-memory storage. With this PR, the server and the client are clearly separated, and only `SimpleMessageQueue` derives from `AbstractMessageQueue`, while the server class is just an object. This way, the simple message queue isn't special anymore, and it behaves exactly like the others. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

yindo added the pull-request label 2026-02-16 01:17:24 -05:00

yindo closed this issue

2026-02-16 01:17:24 -05:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: run-llama/llama_deploy#448