Merge branch 'deps-main' into fix/stackit-embedder

Author: a-klos

.devcontainer/Dockerfile

@@ -4,7 +4,7 @@ ARG DEBIAN_FRONTEND=noninteractive
 ARG USER=vscode
 
 RUN DEBIAN_FRONTEND=noninteractive \
-    && apt-get update \ 
+    && apt-get update \
     && apt-get install -y build-essential --no-install-recommends make \
         ca-certificates \
         git \
@@ -27,7 +27,7 @@ RUN DEBIAN_FRONTEND=noninteractive \
 # Python and poetry installation
 USER $USER
 ARG HOME="/home/$USER"
-ARG PYTHON_VERSION=3.11
+ARG PYTHON_VERSION=3.13
 
 ENV PYENV_ROOT="${HOME}/.pyenv"
 ENV PATH="${PYENV_ROOT}/shims:${PYENV_ROOT}/bin:${HOME}/.local/bin:$PATH"
@@ -40,4 +40,4 @@ RUN echo "done 0" \
     && pyenv global ${PYTHON_VERSION} \
     && echo "done 3" \
     && curl -sSL https://install.python-poetry.org | python3 - \
-    && poetry config virtualenvs.in-project true
+    && poetry config virtualenvs.in-project true

README.md

@@ -38,7 +38,7 @@ The template supports multiple LLM (Large Language Model) providers, such as STA
 
 
 ## 1. Getting Started
-A [`Tiltfile`](./Tiltfile) is provided to get you started :rocket:. If Tilt is new for you, and you want to learn more about it, please take a look at the [Tilt guides](https://docs.tilt.dev/tiltfile_authoring).
+A [`Tiltfile`](./Tiltfile) is provided to get you started :rocket:. If Tilt is new for you, and you want to learn more about it, please take a look at the [Tilt guides](https://docs.tilt.dev/tiltfile_authoring.html).
 
 ### 1.1 Components
 

infrastructure/rag/Chart.lock

@@ -1,7 +1,7 @@
 dependencies:
 - name: langfuse
   repository: https://langfuse.github.io/langfuse-k8s
-  version: 0.13.1
+  version: 1.3.4
 - name: qdrant
   repository: https://qdrant.github.io/qdrant-helm
   version: 1.15.0
@@ -14,5 +14,5 @@ dependencies:
 - name: ollama
   repository: https://otwld.github.io/ollama-helm/
   version: 1.24.0
-digest: sha256:5d274ff927bf767fd41057275293a1751fd0f7834d7499d2b3345eb0f433335b
-generated: "2025-07-31T08:32:30.614652929Z"
+digest: sha256:30cc964388864bd738eb0668367576dbad45eb9d6a885cdc48ef7c07c9692aa0
+generated: "2025-08-04T09:23:39.885193+02:00"

infrastructure/rag/Chart.yaml

@@ -10,7 +10,7 @@ appVersion: "1.0.0"
 dependencies:
 - name: langfuse
   repository: https://langfuse.github.io/langfuse-k8s
-  version: "0.13.1"
+  version: "1.3.4"
   condition: features.langfuse.enabled
 - name: qdrant
   version: 1.15.0
@@ -21,6 +21,7 @@ dependencies:
   version: "15.0.7"
   condition: features.minio.enabled
 - name: keydb
+  alias: keydb
   repository: https://enapter.github.io/charts/
   version: "0.48.0"
   condition: features.keydb.enabled

infrastructure/rag/values.yaml

@@ -477,6 +477,7 @@ langfuse:
         repository: langfuse/langfuse-worker
         tag: "3.88.1"
         pullPolicy: Always
+      port: 3030
 
     # NextAuth configuration
     nextauth:
@@ -548,6 +549,22 @@ langfuse:
       url: "clickhouse://rag-clickhouse:9000"
       ssl: false
       autoMigrate: true
+    resources:
+      limits:
+        cpu: "2"
+        memory: "8Gi"
+      requests:
+        cpu: "2"
+        memory: "4Gi"
+
+    zookeeper:
+      resources:
+        limits:
+          cpu: "2"
+          memory: "2Gi"
+        requests:
+          cpu: "1"
+          memory: "1Gi"
 
   # S3/MinIO Configuration (external MinIO)
   s3:
@@ -588,10 +605,6 @@ minio:
     enabled: false
   mode: standalone
 
-
-
-
-
 ollama:
   image:
     tag: 0.5.1
@@ -607,3 +620,8 @@ ollama:
 qdrant:
   image:
     tag: v1.14.1
+
+keydb:
+  multiMaster: "no"
+  activeReplicas: "no"
+  nodes: 1

libs/README.md

@@ -65,7 +65,7 @@
 from rag_core_lib.impl.settings.ollama_llm_settings import OllamaSettings
 from rag_core_lib.impl.settings.rag_class_types_settings import RAGClassTypeSettings
 from rag_core_lib.impl.settings.stackit_vllm_settings import StackitVllmSettings
-from rag_core_lib.impl.tracers.langfuse_traced_chain import LangfuseTracedGraph
+from rag_core_lib.impl.tracers.langfuse_traced_runnable import LangfuseTracedRunnable
 from rag_core_lib.impl.utils.async_threadsafe_semaphore import AsyncThreadsafeSemaphore
 
 
@@ -138,19 +138,19 @@ class DependencyContainer(DeclarativeContainer):
         chunker=summary_text_splitter,
         semaphore=Singleton(AsyncThreadsafeSemaphore, summarizer_settings.maximum_concurrreny),
     )
+    traced_summarizer = Singleton(
+        LangfuseTracedRunnable,
+        inner_chain=summarizer,
+        settings=langfuse_settings,
+    )
 
     summary_enhancer = List(
-        Singleton(PageSummaryEnhancer, summarizer, chunker_settings),
+        Singleton(PageSummaryEnhancer, traced_summarizer, chunker_settings),
     )
-    untraced_information_enhancer = Singleton(
+    information_enhancer = Singleton(
         GeneralEnhancer,
         summary_enhancer,
     )
-    information_enhancer = Singleton(
-        LangfuseTracedGraph,
-        inner_chain=untraced_information_enhancer,
-        settings=langfuse_settings,
-    )
 
     document_deleter = Singleton(
         DefaultDocumentDeleter, rag_api=rag_api, file_service=file_service, key_value_store=key_value_store

libs/admin-api-lib/src/admin_api_lib/dependency_container.py

@@ -1,5 +1,6 @@
 """Module for the LangchainSummarizer class."""
 
+import asyncio
 import logging
 import traceback
 from typing import Optional
@@ -27,6 +28,8 @@ class LangchainSummarizer(Summarizer):
     document and retries the summarization process if an error occurs.
     """
 
+    RETRY_WAIT_TIME = 10
+
     def __init__(
         self,
         langfuse_manager: LangfuseManager,
@@ -84,7 +87,12 @@ async def ainvoke(self, query: SummarizerInput, config: Optional[RunnableConfig]
                 except Exception as e:
                     logger.error("Error in summarizing langchain doc: %s %s", e, traceback.format_exc())
                     config["tries_remaining"] = tries_remaining - 1
-                    result = await self._create_chain().ainvoke({"text": langchain_document.page_content}, config)
+                    if "rate limit" in str(e).lower() or "ratelimit" in str(e).lower():
+                        logger.warning(
+                            "Rate limit encountered, waiting %d seconds before retry...", self.RETRY_WAIT_TIME
+                        )
+                        await asyncio.sleep(self.RETRY_WAIT_TIME)
+                    result = await self.ainvoke(query, config)
                     # Extract content from AIMessage if it's not already a string
                     content = result.content if hasattr(result, "content") else str(result)
                     outputs.append(content)

libs/admin-api-lib/src/admin_api_lib/impl/summarizer/langchain_summarizer.py

@@ -6,13 +6,13 @@
 from langchain_core.documents import Document
 from langchain_core.runnables import RunnableConfig
 
-from rag_core_lib.chains.async_chain import AsyncChain
+from rag_core_lib.runnables.async_runnable import AsyncRunnable
 
 RetrieverInput = list[Document]
 RetrieverOutput = list[Document]
 
 
-class InformationEnhancer(AsyncChain[RetrieverInput, RetrieverOutput], ABC):
+class InformationEnhancer(AsyncRunnable[RetrieverInput, RetrieverOutput], ABC):
     """The base class for an information enhancer."""
 
     @abstractmethod

libs/admin-api-lib/src/admin_api_lib/information_enhancer/information_enhancer.py

@@ -5,13 +5,13 @@
 
 from langchain_core.runnables import RunnableConfig
 
-from rag_core_lib.chains.async_chain import AsyncChain
+from rag_core_lib.runnables.async_runnable import AsyncRunnable
 
 SummarizerInput = str
 SummarizerOutput = str
 
 
-class Summarizer(AsyncChain[SummarizerInput, SummarizerOutput], ABC):
+class Summarizer(AsyncRunnable[SummarizerInput, SummarizerOutput], ABC):
     """Baseclass for summarizers."""
 
     @abstractmethod