Skip to content

[https://nvbugs/5495789][feat] Optionally disable server GC and worker GC #7995

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

Merged
yuantailing merged 2 commits into from
Sep 26, 2025
Merged

[https://nvbugs/5495789][feat] Optionally disable server GC and worker GC #7995

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9118b8e
Optionally disable server GC and worker GC
yuantailing Sep 25, 2025
203cefb
Cancel int conversion
yuantailing Sep 26, 2025
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
6 changes: 5 additions & 1 deletion tensorrt_llm/commands/serve.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -182,6 +182,10 @@ def launch_server(host: str,
server_role=server_role,
metadata_server_cfg=metadata_server_cfg)

# Optionally disable GC (default: not disabled)
if os.getenv("TRTLLM_SERVER_DISABLE_GC", "0") == "1":
gc.disable()

Comment thread
pcastonguay marked this conversation as resolved.
asyncio.run(server(host, port))


Expand Down Expand Up @@ -518,7 +522,7 @@ def disaggregated(config_file: Optional[str],
# increment, and observed that `count0` (obtained by `gc.get_count()`)
# increases by fewer than 1,000 after every 200,000 requests, while the
# maximum value of `count0` exceeded 3,000,000 during the test.
if int(os.getenv("TRTLLM_DISAGG_SERVER_DISABLE_GC", "1")):
if os.getenv("TRTLLM_DISAGG_SERVER_DISABLE_GC", "1") == "1":
gc.disable()

asyncio.run(server(disagg_cfg.hostname, disagg_cfg.port))
Expand Down
5 changes: 5 additions & 0 deletions tensorrt_llm/executor/worker.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
import gc
import json
import os
import time
Expand Down Expand Up @@ -413,6 +414,10 @@ def notify_proxy_threads_to_quit():
logger.error("Failed to deliver error message to proxy")
return

# Optionally disable GC (default: not disabled)
if os.getenv("TRTLLM_WORKER_DISABLE_GC", "0") == "1":
gc.disable()

with worker:
try:
worker.block_subordinates()
Expand Down