apunkt/plano
1
0
Fork 0
mirror of https://github.com/katanemo/plano.git synced 2026-06-17 15:25:17 +02:00
Code Issues Projects Releases Packages Wiki Activity

fix more

Browse source
This commit is contained in:
Adil Hafeez 2025-02-12 14:48:23 -08:00
parent d2ad943f63
commit 9cb04756c5
No known key found for this signature in database
GPG key ID: 9B18EF7691369645
13 changed files with 181 additions and 90 deletions
Download patch file Download diff file Expand all files Collapse all files

18
arch/tools/cli/config_generator.py
View file

@ -104,7 +104,25 @@ def validate_and_render_schema():
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", {})
if prompt_gateway_listener.get("port") == None:
prompt_gateway_listener["port"] = 10000 # default port for prompt gateway
if prompt_gateway_listener.get("address") == None:
prompt_gateway_listener["address"] = "127.0.0.1"
if prompt_gateway_listener.get("timeout") == None:
prompt_gateway_listener["timeout"] = "10s"
llm_gateway_listener = config_yaml.get("listeners", {}).get("llm_gateway", {})
if llm_gateway_listener.get("port") == None:
llm_gateway_listener["port"] = 12000 # default port for llm gateway
if llm_gateway_listener.get("address") == None:
llm_gateway_listener["address"] = "127.0.0.1"
if llm_gateway_listener.get("timeout") == None:
llm_gateway_listener["timeout"] = "10s"
data = {
"prompt_gateway_listener": prompt_gateway_listener,
"llm_gateway_listener": llm_gateway_listener,
"arch_config": arch_config_string,
"arch_llm_config": arch_llm_config_string,
"arch_clusters": inferred_clusters,

36
arch/tools/cli/core.py
View file

@ -15,12 +15,15 @@ from cli.consts import (
)
from huggingface_hub import snapshot_download
from dotenv import dotenv_values
import yaml
log = getLogger(__name__)
def start_archgw_docker(client, arch_config_file, env):
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)
@ -29,10 +32,10 @@ def start_archgw_docker(client, arch_config_file, env):
image=ARCHGW_DOCKER_IMAGE,
detach=True, # Run in detached mode
ports={
"10000/tcp": 10000,
f"{prompt_gateway_port}/tcp": prompt_gateway_port,
"10001/tcp": 10001,
"11000/tcp": 11000,
"12000/tcp": 12000,
f"{llm_gateway_port}/tcp": llm_gateway_port,
"9901/tcp": 19901,
},
volumes={
@ -50,7 +53,12 @@ def start_archgw_docker(client, arch_config_file, env):
},
extra_hosts={"host.docker.internal": "host-gateway"},
healthcheck={
"test": ["CMD", "curl", "-f", "http://localhost:10000/healthz"],
"test": [
"CMD",
"curl",
"-f",
f"http://localhost:{prompt_gateway_port}/healthz",
],
"interval": 5000000000, # 5 seconds
"timeout": 1000000000, # 1 seconds
"retries": 3,
@ -128,7 +136,25 @@ def start_arch(arch_config_file, env, log_timeout=120, foreground=False):
except docker.errors.NotFound as e:
pass
container = start_archgw_docker(client, arch_config_file, env)
# 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
)
start_time = time.time()