apunkt/plano
1
0
Fork 0
mirror of https://github.com/katanemo/plano.git synced 2026-07-20 16:41:04 +02:00
Code Issues Projects Releases Packages Wiki Activity

Merge branch 'main' into adil/add_acm_demo

Browse source
This commit is contained in:
Adil Hafeez 2025-04-15 15:12:46 -07:00
commit 6edad0870b
No known key found for this signature in database
GPG key ID: 9B18EF7691369645
125 changed files with 6680 additions and 2314 deletions
Download patch file Download diff file Expand all files Collapse all files

3
arch/Dockerfile
View file

@ -28,4 +28,5 @@ COPY arch/arch_config_schema.yaml .
RUN pip install requests
RUN touch /var/log/envoy.log
ENTRYPOINT ["sh","-c", "python config_generator.py && envsubst < /etc/envoy/envoy.yaml > /etc/envoy.env_sub.yaml && envoy -c /etc/envoy.env_sub.yaml --component-log-level wasm:debug 2>&1 | tee /var/log/envoy.log"]
# ENTRYPOINT ["sh","-c", "python config_generator.py && envsubst < /etc/envoy/envoy.yaml > /etc/envoy.env_sub.yaml && envoy -c /etc/envoy.env_sub.yaml --log-level trace 2>&1 | tee /var/log/envoy.log"]
ENTRYPOINT ["sh","-c", "python config_generator.py && envsubst < /etc/envoy/envoy.yaml > /etc/envoy.env_sub.yaml && envoy -c /etc/envoy.env_sub.yaml --component-log-level wasm:info 2>&1 | tee /var/log/envoy.log"]

25
arch/arch_config_schema.yaml
View file

@ -5,8 +5,9 @@ properties:
type: string
listeners:
type: object
additionalProperties: false
properties:
prompt_gateway:
ingress_traffic:
type: object
properties:
address:
@ -20,7 +21,7 @@ properties:
timeout:
type: string
additionalProperties: false
llm_gateway:
egress_traffic:
type: object
properties:
address:
@ -31,7 +32,6 @@ properties:
type: string
enum:
- openai
- huggingface
timeout:
type: string
additionalProperties: false
@ -62,7 +62,7 @@ properties:
properties:
name:
type: string
# this field is deprecated, use provider_interface instead
# provider field is deprecated, use provider_interface instead
provider:
type: string
enum:
@ -78,8 +78,11 @@ properties:
type: string
default:
type: boolean
# endpoint field is deprecated, use base_url instead
endpoint:
type: string
base_url:
type: string
protocol:
type: string
enum:
@ -90,7 +93,6 @@ properties:
additionalProperties: false
required:
- name
- model
overrides:
type: object
properties:
@ -98,6 +100,8 @@ properties:
type: number
optimize_context_window:
type: boolean
use_agent_orchestrator:
type: boolean
system_prompt:
type: string
prompt_targets:
@ -124,7 +128,10 @@ properties:
required:
type: boolean
default:
type: string
anyOf:
- type: string
- type: integer
- type: boolean
description:
type: string
type:
@ -132,7 +139,10 @@ properties:
enum:
type: array
items:
type: string
anyOf:
- type: string
- type: integer
- type: boolean
in_path:
type: boolean
format:
@ -241,5 +251,4 @@ properties:
additionalProperties: false
required:
- version
- listeners
- llm_providers

41
arch/envoy.template.yaml
View file

@ -29,7 +29,7 @@ stats_config:
- 180000
static_resources:
listeners:
- name: arch_listener_http
- name: ingress_traffic
address:
socket_address:
address: {{ prompt_gateway_listener.address }}
@ -55,7 +55,7 @@ static_resources:
random_sampling:
value: {{ arch_tracing.random_sampling }}
{% endif %}
stat_prefix: arch_listener_http
stat_prefix: ingress_traffic
codec_type: AUTO
scheme_header_transformation:
scheme_to_overwrite: https
@ -82,7 +82,7 @@ static_resources:
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
- name: arch_prompt_gateway_listener
- name: ingress_traffic_prompt
address:
socket_address:
address: 0.0.0.0
@ -104,11 +104,11 @@ static_resources:
envoy_grpc:
cluster_name: opentelemetry_collector
timeout: 0.250s
service_name: prompt_processor
service_name: ingress_traffic
random_sampling:
value: {{ arch_tracing.random_sampling }}
{% endif %}
stat_prefix: arch_prompt_gateway_listener
stat_prefix: ingress_traffic
codec_type: AUTO
scheme_header_transformation:
scheme_to_overwrite: https
@ -142,6 +142,19 @@ static_resources:
cluster: {{ llm_cluster_name }}
timeout: 60s
{% endfor %}
{% if agent_orchestrator %}
- match:
prefix: "/"
headers:
- name: "x-arch-llm-provider"
string_match:
exact: {{ agent_orchestrator }}
route:
auto_host_rewrite: true
cluster: {{ agent_orchestrator }}
timeout: 60s
{% endif %}
http_filters:
- name: envoy.filters.http.compressor
typed_config:
@ -201,7 +214,7 @@ static_resources:
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
- name: arch_internal
- name: egress_api_traffic
address:
socket_address:
address: 0.0.0.0
@ -223,11 +236,11 @@ static_resources:
envoy_grpc:
cluster_name: opentelemetry_collector
timeout: 0.250s
service_name: prompt_processor
service_name: egress_api_traffic
random_sampling:
value: {{ arch_tracing.random_sampling }}
{% endif %}
stat_prefix: arch_internal
stat_prefix: egress_api_traffic
codec_type: AUTO
scheme_header_transformation:
scheme_to_overwrite: https
@ -273,12 +286,12 @@ static_resources:
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
- name: arch_listener_http_llm
- name: egress_traffic
address:
socket_address:
address: {{ llm_gateway_listener.address }}
port_value: {{ llm_gateway_listener.port }}
traffic_direction: INBOUND
traffic_direction: OUTBOUND
filter_chains:
- filters:
- name: envoy.filters.network.http_connection_manager
@ -299,7 +312,7 @@ static_resources:
random_sampling:
value: {{ arch_tracing.random_sampling }}
{% endif %}
stat_prefix: arch_listener_http
stat_prefix: egress_traffic
codec_type: AUTO
scheme_header_transformation:
scheme_to_overwrite: https
@ -326,7 +339,7 @@ static_resources:
typed_config:
"@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
- name: arch_listener_llm
- name: egress_traffic_llm
address:
socket_address:
address: 0.0.0.0
@ -347,11 +360,11 @@ static_resources:
envoy_grpc:
cluster_name: opentelemetry_collector
timeout: 0.250s
service_name: llm_gateway
service_name: egress_traffic_llm
random_sampling:
value: {{ arch_tracing.random_sampling }}
{% endif %}
stat_prefix: arch_listener_http
stat_prefix: egress_traffic
codec_type: AUTO
scheme_header_transformation:
scheme_to_overwrite: https

2
arch/tools/README.md
View file

@ -19,7 +19,7 @@ source venv/bin/activate
### Step 3: Run the build script
```bash
pip install archgw==0.2.1
pip install archgw==0.2.6
```
## Uninstall Instructions: archgw CLI

57
arch/tools/cli/config_generator.py
View file

@ -3,6 +3,7 @@ import os
from jinja2 import Environment, FileSystemLoader
import yaml
from jsonschema import validate
from urllib.parse import urlparse
ENVOY_CONFIG_TEMPLATE_FILE = os.getenv(
"ENVOY_CONFIG_TEMPLATE_FILE", "envoy.template.yaml"
@ -47,7 +48,7 @@ def validate_and_render_schema():
arch_config_schema = file.read()
config_yaml = yaml.safe_load(arch_config)
config_schema_yaml = yaml.safe_load(arch_config_schema)
_ = yaml.safe_load(arch_config_schema)
inferred_clusters = {}
endpoints = config_yaml.get("endpoints", {})
@ -91,6 +92,9 @@ def validate_and_render_schema():
del llm_provider["provider"]
updated_llm_providers.append(llm_provider)
if llm_provider.get("endpoint") and llm_provider.get("base_url"):
raise Exception("Please provide either endpoint or base_url, not both")
if llm_provider.get("endpoint", None):
endpoint = llm_provider["endpoint"]
protocol = llm_provider.get("protocol", "http")
@ -98,13 +102,39 @@ def validate_and_render_schema():
endpoint, protocol
)
llms_with_endpoint.append(llm_provider)
elif llm_provider.get("base_url", None):
base_url = llm_provider["base_url"]
urlparse_result = urlparse(base_url)
if llm_provider.get("port"):
raise Exception("Please provider port in base_url")
if urlparse_result.scheme == "" or urlparse_result.scheme not in [
"http",
"https",
]:
raise Exception(
"Please provide a valid URL with scheme (http/https) in base_url"
)
protocol = urlparse_result.scheme
port = urlparse_result.port
if port is None:
if protocol == "http":
port = 80
else:
port = 443
endpoint = urlparse_result.hostname
llm_provider["endpoint"] = endpoint
llm_provider["port"] = port
llm_provider["protocol"] = protocol
llms_with_endpoint.append(llm_provider)
config_yaml["llm_providers"] = updated_llm_providers
arch_config_string = yaml.dump(config_yaml)
arch_llm_config_string = yaml.dump(config_yaml)
prompt_gateway_listener = config_yaml.get("listeners", {}).get("prompt_gateway", {})
prompt_gateway_listener = config_yaml.get("listeners", {}).get(
"ingress_traffic", {}
)
if prompt_gateway_listener.get("port") == None:
prompt_gateway_listener["port"] = 10000 # default port for prompt gateway
if prompt_gateway_listener.get("address") == None:
@ -112,7 +142,7 @@ def validate_and_render_schema():
if prompt_gateway_listener.get("timeout") == None:
prompt_gateway_listener["timeout"] = "10s"
llm_gateway_listener = config_yaml.get("listeners", {}).get("llm_gateway", {})
llm_gateway_listener = config_yaml.get("listeners", {}).get("egress_traffic", {})
if llm_gateway_listener.get("port") == None:
llm_gateway_listener["port"] = 12000 # default port for llm gateway
if llm_gateway_listener.get("address") == None:
@ -120,6 +150,26 @@ def validate_and_render_schema():
if llm_gateway_listener.get("timeout") == None:
llm_gateway_listener["timeout"] = "10s"
use_agent_orchestrator = config_yaml.get("overrides", {}).get(
"use_agent_orchestrator", False
)
agent_orchestrator = None
if use_agent_orchestrator:
print("Using agent orchestrator")
if len(endpoints) == 0:
raise Exception(
"Please provide agent orchestrator in the endpoints section in your arch_config.yaml file"
)
elif len(endpoints) > 1:
raise Exception(
"Please provide single agent orchestrator in the endpoints section in your arch_config.yaml file"
)
else:
agent_orchestrator = list(endpoints.keys())[0]
print("agent_orchestrator: ", agent_orchestrator)
data = {
"prompt_gateway_listener": prompt_gateway_listener,
"llm_gateway_listener": llm_gateway_listener,
@ -129,6 +179,7 @@ def validate_and_render_schema():
"arch_llm_providers": config_yaml["llm_providers"],
"arch_tracing": arch_tracing,
"local_llms": llms_with_endpoint,
"agent_orchestrator": agent_orchestrator,
}
rendered = template.render(data)

200
arch/tools/cli/core.py
View file

@ -2,110 +2,49 @@ import subprocess
import os
import time
import sys
import glob
import docker
from docker.errors import DockerException
from cli.utils import getLogger, update_docker_host_env
import yaml
from cli.utils import getLogger
from cli.consts import (
ARCHGW_DOCKER_IMAGE,
ARCHGW_DOCKER_NAME,
KATANEMO_LOCAL_MODEL_LIST,
MODEL_SERVER_LOG_FILE,
ACCESS_LOG_FILES,
)
from huggingface_hub import snapshot_download
from dotenv import dotenv_values
import yaml
import subprocess
from cli.docker_cli import (
docker_container_status,
docker_remove_container,
docker_start_archgw_detached,
docker_stop_container,
health_check_endpoint,
stream_gateway_logs,
)
log = getLogger(__name__)
def start_archgw_docker(
client, arch_config_file, env, prompt_gateway_port, llm_gateway_port
):
logs_path = "~/archgw_logs"
logs_path_abs = os.path.expanduser(logs_path)
def _get_gateway_ports(arch_config_file: str) -> tuple:
PROMPT_GATEWAY_DEFAULT_PORT = 10000
LLM_GATEWAY_DEFAULT_PORT = 12000
return client.containers.run(
name=ARCHGW_DOCKER_NAME,
image=ARCHGW_DOCKER_IMAGE,
detach=True, # Run in detached mode
ports={
f"{prompt_gateway_port}/tcp": prompt_gateway_port,
"10001/tcp": 10001,
"11000/tcp": 11000,
f"{llm_gateway_port}/tcp": llm_gateway_port,
"9901/tcp": 19901,
},
volumes={
f"{arch_config_file}": {
"bind": "/app/arch_config.yaml",
"mode": "ro",
},
"/etc/ssl/cert.pem": {"bind": "/etc/ssl/cert.pem", "mode": "ro"},
logs_path_abs: {"bind": "/var/log"},
},
environment={
"OTEL_TRACING_HTTP_ENDPOINT": "http://host.docker.internal:4318/v1/traces",
"MODEL_SERVER_PORT": os.getenv("MODEL_SERVER_PORT", "51000"),
**env,
},
extra_hosts={"host.docker.internal": "host-gateway"},
healthcheck={
"test": [
"CMD",
"curl",
"-f",
f"http://localhost:{prompt_gateway_port}/healthz",
],
"interval": 5000000000, # 5 seconds
"timeout": 1000000000, # 1 seconds
"retries": 3,
},
# parse arch_config_file yaml file and get prompt_gateway_port
arch_config_dict = {}
with open(arch_config_file) as f:
arch_config_dict = yaml.safe_load(f)
prompt_gateway_port = (
arch_config_dict.get("listeners", {})
.get("ingress_traffic", {})
.get("port", PROMPT_GATEWAY_DEFAULT_PORT)
)
llm_gateway_port = (
arch_config_dict.get("listeners", {})
.get("egress_traffic", {})
.get("port", LLM_GATEWAY_DEFAULT_PORT)
)
def stream_gateway_logs(follow):
"""
Stream logs from the arch gateway service.
"""
log.info("Logs from arch gateway service.")
options = ["docker", "logs", "archgw"]
if follow:
options.append("-f")
try:
# Run `docker-compose logs` to stream logs from the gateway service
subprocess.run(
options,
check=True,
stdout=sys.stdout,
stderr=sys.stderr,
)
except subprocess.CalledProcessError as e:
log.info(f"Failed to stream logs: {str(e)}")
def stream_access_logs(follow):
"""
Get the archgw access logs
"""
log_file_pattern_expanded = os.path.expanduser(ACCESS_LOG_FILES)
log_files = glob.glob(log_file_pattern_expanded)
stream_command = ["tail"]
if follow:
stream_command.append("-f")
stream_command.extend(log_files)
subprocess.run(
stream_command,
check=True,
stdout=sys.stdout,
stderr=sys.stderr,
)
return prompt_gateway_port, llm_gateway_port
def start_arch(arch_config_file, env, log_timeout=120, foreground=False):
@ -119,73 +58,58 @@ def start_arch(arch_config_file, env, log_timeout=120, foreground=False):
log.info("Starting arch gateway")
try:
try:
client = docker.from_env()
except DockerException as e:
# try setting up the docker host environment variable and retry
update_docker_host_env()
client = docker.from_env()
archgw_container_status = docker_container_status(ARCHGW_DOCKER_NAME)
if archgw_container_status != "not found":
log.info("archgw found in docker, stopping and removing it")
docker_stop_container(ARCHGW_DOCKER_NAME)
docker_remove_container(ARCHGW_DOCKER_NAME)
try:
container = client.containers.get("archgw")
log.info("archgw container found in docker, stopping and removing it")
# ensure that previous docker container is stopped and removed
container.stop()
container.remove()
log.info("Stopped and removed archgw container")
except docker.errors.NotFound as e:
pass
prompt_gateway_port, llm_gateway_port = _get_gateway_ports(arch_config_file)
# parse arch_config_file yaml file and get prompt_gateway_port
arch_config_dict = {}
with open(arch_config_file) as f:
arch_config_dict = yaml.safe_load(f)
prompt_gateway_port = (
arch_config_dict.get("listeners", {})
.get("prompt_gateway", {})
.get("port", 10000)
)
llm_gateway_port = (
arch_config_dict.get("listeners", {})
.get("llm_gateway", {})
.get("port", 12000)
)
container = start_archgw_docker(
client, arch_config_file, env, prompt_gateway_port, llm_gateway_port
return_code, _, archgw_stderr = docker_start_archgw_detached(
arch_config_file,
os.path.expanduser("~/archgw_logs"),
env,
prompt_gateway_port,
llm_gateway_port,
)
if return_code != 0:
log.info("Failed to start arch gateway: " + str(return_code))
log.info("stderr: " + archgw_stderr)
sys.exit(1)
start_time = time.time()
while True:
container = client.containers.get(container.id)
prompt_gateway_health_check_status = health_check_endpoint(
f"http://localhost:{prompt_gateway_port}/healthz"
)
llm_gateway_health_check_status = health_check_endpoint(
f"http://localhost:{llm_gateway_port}/healthz"
)
archgw_status = docker_container_status(ARCHGW_DOCKER_NAME)
current_time = time.time()
elapsed_time = current_time - start_time
# Check if timeout is reached
if elapsed_time > log_timeout:
log.info(f"Stopping log monitoring after {log_timeout} seconds.")
log.info(f"stopping log monitoring after {log_timeout} seconds.")
break
container_status = container.attrs["State"]["Health"]["Status"]
if container_status == "healthy":
log.info("Container is healthy!")
if prompt_gateway_health_check_status or llm_gateway_health_check_status:
log.info("archgw is running and is healthy!")
break
else:
log.info(f"Container health status: {container_status}")
log.info(f"archgw status: {archgw_status}, health status: starting")
time.sleep(1)
if foreground:
for line in container.logs(stream=True):
print(line.decode("utf-8").strip("\n"))
stream_gateway_logs(follow=True)
except KeyboardInterrupt:
log.info("Keyboard interrupt received, stopping arch gateway service.")
stop_arch()
except docker.errors.APIError as e:
log.info(f"Failed to start Arch: {str(e)}")
def stop_arch():
@ -199,10 +123,10 @@ def stop_arch():
try:
subprocess.run(
["docker", "stop", "archgw"],
["docker", "stop", ARCHGW_DOCKER_NAME],
)
subprocess.run(
["docker", "remove", "archgw"],
["docker", "rm", ARCHGW_DOCKER_NAME],
)
log.info("Successfully shut down arch gateway service.")

133
arch/tools/cli/docker_cli.py Normal file
View file

@ -0,0 +1,133 @@
import subprocess
import json
import sys
import requests
from cli.consts import ARCHGW_DOCKER_IMAGE, ARCHGW_DOCKER_NAME
from cli.utils import getLogger
log = getLogger(__name__)
def docker_container_status(container: str) -> str:
result = subprocess.run(
["docker", "inspect", "--type=container", container],
capture_output=True,
text=True,
check=False,
)
if result.returncode != 0:
return "not found"
container_status = json.loads(result.stdout)[0]
return container_status.get("State", {}).get("Status", "")
def docker_stop_container(container: str) -> str:
result = subprocess.run(
["docker", "stop", container], capture_output=True, text=True, check=False
)
return result.returncode
def docker_remove_container(container: str) -> str:
result = subprocess.run(
["docker", "rm", container], capture_output=True, text=True, check=False
)
return result.returncode
def docker_start_archgw_detached(
arch_config_file: str,
logs_path_abs: str,
env: dict,
prompt_gateway_port,
llm_gateway_port,
) -> str:
env_args = [item for key, value in env.items() for item in ["-e", f"{key}={value}"]]
port_mappings = [
f"{prompt_gateway_port}:{prompt_gateway_port}",
f"{llm_gateway_port}:{llm_gateway_port}",
"9901:19901",
]
port_mappings_args = [item for port in port_mappings for item in ("-p", port)]
volume_mappings = [
f"{logs_path_abs}:/var/log:rw",
f"{arch_config_file}:/app/arch_config.yaml:ro",
# "/Users/adilhafeez/src/intelligent-prompt-gateway/crates/target/wasm32-wasip1/release:/etc/envoy/proxy-wasm-plugins:ro",
]
volume_mappings_args = [
item for volume in volume_mappings for item in ("-v", volume)
]
options = [
"docker",
"run",
"-d",
"--name",
ARCHGW_DOCKER_NAME,
*port_mappings_args,
*volume_mappings_args,
*env_args,
"--add-host",
"host.docker.internal:host-gateway",
ARCHGW_DOCKER_IMAGE,
]
result = subprocess.run(options, capture_output=True, text=True, check=False)
return result.returncode, result.stdout, result.stderr
def health_check_endpoint(endpoint: str) -> bool:
try:
response = requests.get(endpoint)
if response.status_code == 200:
return True
except requests.RequestException as e:
pass
return False
def stream_gateway_logs(follow):
"""
Stream logs from the arch gateway service.
"""
log.info("Logs from arch gateway service.")
options = ["docker", "logs"]
if follow:
options.append("-f")
options.append(ARCHGW_DOCKER_NAME)
try:
# Run `docker-compose logs` to stream logs from the gateway service
subprocess.run(
options,
check=True,
stdout=sys.stdout,
stderr=sys.stderr,
)
except subprocess.CalledProcessError as e:
log.info(f"Failed to stream logs: {str(e)}")
def docker_validate_archgw_schema(arch_config_file):
result = subprocess.run(
[
"docker",
"run",
"--rm",
"-v",
f"{arch_config_file}:/app/arch_config.yaml:ro",
"--entrypoint",
"python",
ARCHGW_DOCKER_IMAGE,
"config_generator.py",
],
capture_output=True,
text=True,
check=False,
)
return result.returncode, result.stdout, result.stderr

25
arch/tools/cli/main.py
View file

@ -5,11 +5,12 @@ import subprocess
import multiprocessing
import importlib.metadata
from cli import targets
from cli.docker_cli import docker_validate_archgw_schema, stream_gateway_logs
from cli.utils import (
getLogger,
get_llm_provider_access_keys,
load_env_file_to_dict,
validate_schema,
stream_access_logs,
)
from cli.core import (
start_arch_modelserver,
@ -17,12 +18,9 @@ from cli.core import (
start_arch,
stop_arch,
download_models_from_hf,
stream_access_logs,
stream_gateway_logs,
)
from cli.consts import (
KATANEMO_DOCKERHUB_REPO,
KATANEMO_LOCAL_MODEL_LIST,
SERVICE_NAME_ARCHGW,
SERVICE_NAME_MODEL_SERVER,
SERVICE_ALL,
@ -174,17 +172,24 @@ def up(file, path, service, foreground):
log.info(f"Validating {arch_config_file}")
try:
validate_schema(arch_config_file)
except Exception as e:
log.info(f"Exiting archgw up: validation failed")
log.info(f"Error: {str(e)}")
(
validation_return_code,
validation_stdout,
validation_stderr,
) = docker_validate_archgw_schema(arch_config_file)
if validation_return_code != 0:
log.info(f"Error: Validation failed. Exiting")
log.info(f"Validation stdout: {validation_stdout}")
log.info(f"Validation stderr: {validation_stderr}")
sys.exit(1)
log.info("Starting arch model server and arch gateway")
# Set the ARCH_CONFIG_FILE environment variable
env_stage = {}
env_stage = {
"OTEL_TRACING_HTTP_ENDPOINT": "http://host.docker.internal:4318/v1/traces",
"MODEL_SERVER_PORT": os.getenv("MODEL_SERVER_PORT", "51000"),
}
env = os.environ.copy()
# check if access_keys are preesnt in the config file
access_keys = get_llm_provider_access_keys(arch_config_file=arch_config_file)

1
arch/tools/cli/targets.py
View file

@ -2,7 +2,6 @@ import ast
import sys
import yaml
from typing import Any
from pydantic import BaseModel
FLASK_ROUTE_DECORATORS = ["route", "get", "post", "put", "delete", "patch"]
FASTAPI_ROUTE_DECORATORS = ["get", "post", "put", "delete", "patch"]

84
arch/tools/cli/utils.py
View file

@ -1,10 +1,11 @@
import glob
import os
import subprocess
import sys
import yaml
import logging
import docker
from docker.errors import DockerException
from cli.consts import ARCHGW_DOCKER_IMAGE, ARCHGW_DOCKER_NAME
from cli.consts import ACCESS_LOG_FILES
logging.basicConfig(
level=logging.INFO,
@ -21,63 +22,6 @@ def getLogger(name="cli"):
log = getLogger(__name__)
def update_docker_host_env():
"""
Update DOCKER_HOST environment variable to use the local Docker socket
"""
if os.getenv("DOCKER_HOST"):
return
default_docker_socket = os.getenv("DEFAULT_DOCKER_SOCKET", "/var/run/docker.sock")
if not os.path.exists(default_docker_socket):
home_dir = os.getenv("HOME")
docker_host = f"unix://{home_dir}/.docker/run/docker.sock"
log.info(
f"Default docker socket {default_docker_socket} not found, using {docker_host}"
)
os.environ["DOCKER_HOST"] = docker_host
def validate_schema(arch_config_file: str) -> None:
try:
try:
client = docker.from_env()
except DockerException as e:
# try setting up the docker host environment variable and retry
update_docker_host_env()
client = docker.from_env()
container = client.containers.run(
image=ARCHGW_DOCKER_IMAGE,
volumes={
f"{arch_config_file}": {
"bind": "/app/arch_config.yaml",
"mode": "ro",
},
},
entrypoint=["python", "config_generator.py"],
detach=True,
)
# Wait for the container to finish and get the exit code
exit_code = container.wait()
# Check exit code for validation success
if exit_code["StatusCode"] != 0:
# Validation failed (non-zero exit code)
logs = container.logs().decode() # Get container logs for debugging
raise ValueError(
f"Validation failed. Container exited with code {exit_code}.\nLogs:\n{logs}"
)
# Successful validation (exit code 0)
log.info("Schema validation successful!")
except docker.errors.APIError as e:
# Handle container creation error
raise ValueError(f"Failed to create container: {e}")
def get_llm_provider_access_keys(arch_config_file):
with open(arch_config_file, "r") as file:
arch_config = file.read()
@ -127,3 +71,23 @@ def load_env_file_to_dict(file_path):
env_dict[key] = value
return env_dict
def stream_access_logs(follow):
"""
Get the archgw access logs
"""
log_file_pattern_expanded = os.path.expanduser(ACCESS_LOG_FILES)
log_files = glob.glob(log_file_pattern_expanded)
stream_command = ["tail"]
if follow:
stream_command.append("-f")
stream_command.extend(log_files)
subprocess.run(
stream_command,
check=True,
stdout=sys.stdout,
stderr=sys.stderr,
)

2224
arch/tools/poetry.lock generated
View file

File diff suppressed because it is too large Load diff

6
arch/tools/pyproject.toml
View file

@ -1,6 +1,6 @@
[tool.poetry]
name = "archgw"
version = "0.2.1"
version = "0.2.6"
description = "Python-based CLI tool to manage Arch Gateway."
authors = ["Katanemo Labs, Inc."]
packages = [
@ -10,13 +10,11 @@ readme = "README.md"
[tool.poetry.dependencies]
python = "^3.10"
archgw_modelserver = "^0.2.1"
archgw_modelserver = "^0.2.6"
click = "^8.1.7"
jinja2 = "^3.1.4"
jsonschema = "^4.23.0"
setuptools = "75.5.0"
docker = "^7.1.0"
python-dotenv = "^1.0.1"
pyyaml = "^6.0.2"
[tool.poetry.scripts]

20
arch/validate_arch_config.sh Normal file
View file

@ -0,0 +1,20 @@
#!/bin/bash
failed_files=()
for file in $(find . -name arch_config.yaml -o -name arch_config_full_reference.yaml); do
echo "Validating $file..."
if ! docker run --rm -v "$(pwd)/$file:/app/arch_config.yaml:ro" --entrypoint /bin/sh katanemo/archgw:latest -c "python config_generator.py" 2>&1 > /dev/null ; then
echo "Validation failed for $file"
failed_files+=("$file")
fi
done
# Print summary of failed files
if [ ${#failed_files[@]} -ne 0 ]; then
echo -e "\nValidation failed for the following files:"
printf '%s\n' "${failed_files[@]}"
exit 1
else
echo -e "\nAll files validated successfully!"
fi