mfitton

cluster_name: repro-issue-13090

min_workers: 0
max_workers: 0

docker:
    image: anyscale/ray-ml:latest
    container_name: ray_container
    pull_before_run: True

mfitton

+-------------------------------------------+------------+-------+--------+------------------+---------+----------+----------------------+----------------------+--------------------+
| Trial name                                | status     | loc   |   iter |   total time (s) |      ts |   reward |   episode_reward_max |   episode_reward_min |   episode_len_mean |
|-------------------------------------------+------------+-------+--------+------------------+---------+----------+----------------------+----------------------+--------------------|
| A2C_BreakoutNoFrameskip-v4_ed940_00000    | TERMINATED |       |    354 |          3601.99 | 3929000 |  142.56  |              330     |               24     |            5552.71 |
| A2C_BreakoutNoFrameskip-v4_ed940_00001    | TERMINATED |       |    354 |          3609.69 | 3975000 |  164.16  |              396     |               26     |            5796.55 |
| APEX_BreakoutNoFrameskip-v4_ed940_00002   | TERMINATED |       |     92 |          3616.25 | 6282880 |   12

mfitton

####################################################################
# All nodes in this cluster will auto-terminate in 1 hour
####################################################################

# An unique identifier for the head node and workers of this cluster.
cluster_name: autoscaler-stress-test-1.0.0

# The minimum number of workers nodes to launch in addition to the head
# node. This number should be >= 0.
min_workers: 100

mfitton

Iteration 1023:
  - Iteration time: 9.326589822769165.
  - Absolute time: 1607723877.3213854.
  - Total elapsed time: 7786.904662370682.
2020-12-11 21:58:03,954 WARNING services.py:1640 -- WARNING: The object store is using /tmp instead of /dev/shm because /dev/shm has only 67108864 bytes available. This may slow down performance! You may be able to free up space by deleting files in /dev/shm or terminating any running plasma_store_server processes. If you are inside a Docker container, you may need to pass an argument with the flag '--shm-size' to 'docker run'.
(pid=raylet) E1211 21:58:05.133839404   73164 server_chttp2.cc:40]        {"created":"@1607723885.133752839","description":"No address added out of total 1 resolved","file":"external/com_github_grpc_grpc/src/core/ext/transport/chttp2/server/chttp2_server.cc","file_line":394,"referenced_errors":[{"created":"@1607723885.133750775","description":"Failed to add any wildcard listeners","file":"external/com_github_grpc_grpc/src/core/lib/iomgr/tcp_server_pos

mfitton

import ray
import streamlit as st

def fetch_actor():
    actor = retry_until_success(lambda: ray.get_actor("streamlitactor"))
    return actor

def main():
    ray.init(address="auto", ignore_reinit_error=True)
    actor = fetch_actor()

mfitton

import streamlit as st
import time
import session_state
from datetime import datetime, timedelta
from util import get_db_connection
from PIL import Image
import ray

def retry_until_success(f, timeout=15):
    end = datetime.now() + timedelta(seconds=timeout)

mfitton

import uuid

import ray
from ray import serve

from util import ImpressionStore, choose_ensemble_results


@ray.remote
class ComposedModel:

mfitton

import ray
import os
import time
from datetime import datetime, timedelta
import streamlit as st

def main():
    ray.init(address="auto")
    actor_name: Optional[str] = os.getenv("ACTOR_NAME")
    if not actor_name:

mfitton

import ray
import os
import time
from datetime import datetime, timedelta
import streamlit as st

def main():
    ray.init(address="auto")
    actor_name: Optional[str] = os.getenv("ACTOR_NAME")
    if not actor_name:

mfitton

import ray
import os
import time
from datetime import datetime, timedelta
import streamlit as st

def main():
    ray.init(address="auto")
    actor_name: Optional[str] = os.getenv("ACTOR_NAME")
    if not actor_name: