Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add unique log files for each mongo worker #2016

Merged
merged 7 commits into from
Apr 5, 2024
Merged

Add unique log files for each mongo worker #2016

Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
3461d6e
Add log files for each mongo worker
Cmurilochem Mar 20, 2024
09d468b
Include runcard name to mongo workers log filenames
Cmurilochem Mar 22, 2024
a3731fc
Change self.output_path to self.job_output_name
Cmurilochem Mar 29, 2024
4bdcca4
Open mongo-workers using 'with'
Cmurilochem Apr 5, 2024
6269f8c
Update n3fit/src/n3fit/hyper_optimization/mongofiletrials.py
Cmurilochem Apr 5, 2024
8e6d642
Update n3fit/src/n3fit/hyper_optimization/mongofiletrials.py
Cmurilochem Apr 5, 2024
6f71343
Update output_folder_name
Cmurilochem Apr 5, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 10 additions & 5 deletions n3fit/src/n3fit/hyper_optimization/mongofiletrials.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,7 @@
Hyperopt trial object for parallel hyperoptimization with MongoDB.
Data are fetched from MongoDB databases and stored in the form of json and tar.gz files within the nnfit folder.
"""

import json
import logging
import os
Expand Down Expand Up @@ -154,6 +155,7 @@ def __init__(
f"mongo://{self.db_host}:{self.db_port}/{self._process_db_name(self.db_name)}/jobs"
)
self.workers = []
self.output_folder_name = replica_path.parts[-3]

self._store_trial = False
self._json_file = replica_path / "tries.json"
Expand Down Expand Up @@ -195,7 +197,6 @@ def refresh(self):

# write json to disk
if self._store_trial:
log.info("Storing scan in %s", self._json_file)
local_trials = []
for idx, t in enumerate(self._dynamic_trials):
local_trials.append(t)
Expand Down Expand Up @@ -264,10 +265,14 @@ def start_mongo_workers(
# avoid memory fragmentation issues?
# my_env["TF_GPU_ALLOCATOR"] = "cuda_malloc_async"

# run mongo workers
# we could use stdout=subprocess.DEVNULL and stderr=subprocess.DEVNULL in Popen to suppress output info
worker = subprocess.Popen(args, env=my_env)
self.workers.append(worker)
# create log files to redirect the mongo-workers output
mongo_workers_logfile = f"mongo-worker_{i+1}_{self.output_folder_name}.log"
with open(mongo_workers_logfile, mode='w', encoding="utf-8") as log_file:
# run mongo workers
worker = subprocess.Popen(
args, env=my_env, stdout=log_file, stderr=subprocess.STDOUT
)
self.workers.append(worker)
log.info(f"Started mongo worker {i+1}/{self.num_workers}")
except OSError as err:
msg = f"Failed to execute {args}. Make sure you have MongoDB installed."
Expand Down