github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 05:05:37 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #837] Ollama server stops responding. #78078

New Issue
Closed
opened 2026-05-08 21:49:24 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-05-08 21:49:24 -05:00
Owner
Copy Link

Originally created by @UYousafzai on GitHub (Oct 18, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/837

API Call:

/api/generate

Note:

The server keeps working until left idle for a long time, I get this trace with error: "no child processes" upon request after the idle state.

I have also observed this behavior upon requesting from two different terminals at the same time.

Full Trace:

llama_print_timings: load time = 6042.33 ms
llama_print_timings: sample time = 19.14 ms / 54 runs ( 0.35 ms per token, 2822.05 tokens per second)
llama_print_timings: prompt eval time = 0.00 ms / 1 tokens ( 0.00 ms per token, inf tokens per second)
llama_print_timings: eval time = 5461.00 ms / 54 runs ( 101.13 ms per token, 9.89 tokens per second)
llama_print_timings: total time = 5490.18 ms
{"timestamp":1697544483,"level":"INFO","function":"log_server_request","line":1157,"message":"request","remote_addr":"127.0.0.1","remote_port":41622,"status":200,"method":"POST","path":"/completion","params":{}}
{"timestamp":1697544483,"level":"INFO","function":"log_server_request","line":1157,"message":"request","remote_addr":"127.0.0.1","remote_port":53162,"status":200,"method":"POST","path":"/tokenize","params":{}}
[GIN] 2023/10/17 - 12:08:03 | 200 | 5.491869831s | 172.17.0.1 | POST "/api/generate"
2023/10/17 12:13:03 llama.go:323: llama runner exited with error: wait: no child processes

Key Trace:

[GIN] 2023/10/17 - 12:08:03 | 200 | 5.491869831s | 172.17.0.1 | POST "/api/generate"
2023/10/17 12:13:03 llama.go:323: llama runner exited with error: wait: no child processes

Originally created by @UYousafzai on GitHub (Oct 18, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/837 API Call: ---------------- /api/generate Note: ---------------- The server keeps working until left idle for a long time, I get this trace with error: "no child processes" upon request after the idle state. I have also observed this behavior upon requesting from two different terminals at the same time. Full Trace: ---------------- llama_print_timings: load time = 6042.33 ms llama_print_timings: sample time = 19.14 ms / 54 runs ( 0.35 ms per token, 2822.05 tokens per second) llama_print_timings: prompt eval time = 0.00 ms / 1 tokens ( 0.00 ms per token, inf tokens per second) llama_print_timings: eval time = 5461.00 ms / 54 runs ( 101.13 ms per token, 9.89 tokens per second) llama_print_timings: total time = 5490.18 ms {"timestamp":1697544483,"level":"INFO","function":"log_server_request","line":1157,"message":"request","remote_addr":"127.0.0.1","remote_port":41622,"status":200,"method":"POST","path":"/completion","params":{}} {"timestamp":1697544483,"level":"INFO","function":"log_server_request","line":1157,"message":"request","remote_addr":"127.0.0.1","remote_port":53162,"status":200,"method":"POST","path":"/tokenize","params":{}} [GIN] 2023/10/17 - 12:08:03 | 200 | 5.491869831s | 172.17.0.1 | POST "/api/generate" 2023/10/17 12:13:03 llama.go:323: llama runner exited with error: wait: no child processes Key Trace: ---------------- [GIN] 2023/10/17 - 12:08:03 | 200 | 5.491869831s | 172.17.0.1 | POST "/api/generate" 2023/10/17 12:13:03 llama.go:323: llama runner exited with error: wait: no child processes
GiteaMirror added the bug label 2026-05-08 21:49:24 -05:00
GiteaMirror commented 2026-05-08 21:51:56 -05:00
Author
Owner
Copy Link

@mchiang0610 commented on GitHub (Oct 18, 2023):

Hi @UYousafzai thank you for submitting this. May I ask how and where you are running the Ollama server?

<!-- gh-comment-id:1768939776 --> @mchiang0610 commented on GitHub (Oct 18, 2023): Hi @UYousafzai thank you for submitting this. May I ask how and where you are running the Ollama server?
GiteaMirror commented 2026-05-08 21:51:58 -05:00
Author
Owner
Copy Link

@UYousafzai commented on GitHub (Oct 18, 2023):

I am running this on my local server:

###OS Details
Distributor ID: Ubuntu
Description: Ubuntu 22.04.3 LTS
Release: 22.04
Codename: jammy

###Processor Details:
Model Name: AMD Ryzen 9 7950X 16-Core Processor
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit

Note:
I do have a 4090 GPU however for the testing I am nor running the server in GPU mode and I am only utilizing the CPU.

With the short time that I had I tried looking into any specific timeout functions for the server but I couldn't find any specific ones that triggered on being left idle (this is just my assumption that the ones I did find weren't of this nature).

<!-- gh-comment-id:1769097837 --> @UYousafzai commented on GitHub (Oct 18, 2023): I am running this on my local server: ###OS Details Distributor ID: Ubuntu Description: Ubuntu 22.04.3 LTS Release: 22.04 Codename: jammy ###Processor Details: Model Name: AMD Ryzen 9 7950X 16-Core Processor Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Note: I do have a 4090 GPU however for the testing I am nor running the server in GPU mode and I am only utilizing the CPU. With the short time that I had I tried looking into any specific timeout functions for the server but I couldn't find any specific ones that triggered on being left idle (this is just my assumption that the ones I did find weren't of this nature).
GiteaMirror commented 2026-05-08 21:52:00 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Oct 18, 2023):

Other than the log message, is there any indication ollama isn't working as expected? The model gets unloaded after 5 minutes of inactivity to free up resources. The logs indicate this behaviour seems to be working albeit with an unexpected error message

<!-- gh-comment-id:1769453437 --> @mxyng commented on GitHub (Oct 18, 2023): Other than the log message, is there any indication ollama isn't working as expected? The model gets unloaded after 5 minutes of inactivity to free up resources. The logs indicate this behaviour seems to be working albeit with an unexpected error message
GiteaMirror commented 2026-05-08 21:52:01 -05:00
Author
Owner
Copy Link

@UYousafzai commented on GitHub (Oct 19, 2023):

I checked through the code and I was trying to see if I can spot anything but given I barely had an hour to look at this so I couldn't really figure it out.

could you tell me where I can extend the time that has been set for freeing up resources, for the time being I am testing a couple of things and I am okay with resources being held indefinitely.

additionally if you would like to see some logs and you know where exactly they are stored let me know and I can produce them for your debugging purpose.

<!-- gh-comment-id:1770877272 --> @UYousafzai commented on GitHub (Oct 19, 2023): I checked through the code and I was trying to see if I can spot anything but given I barely had an hour to look at this so I couldn't really figure it out. could you tell me where I can extend the time that has been set for freeing up resources, for the time being I am testing a couple of things and I am okay with resources being held indefinitely. additionally if you would like to see some logs and you know where exactly they are stored let me know and I can produce them for your debugging purpose.
GiteaMirror commented 2026-05-08 21:52:02 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Oct 19, 2023):

The 5 minute timeout is not currently user configurable. If you want to keep it alive, you can periodically send an empty generate request. Something like this will work:

while :; do curl localhost:11434/api/generate -d '{"model":"<model>","system":"","prompt":"","template":""}'; sleep 60; done

There are plans on improving the scheduling and session time outs but not at this moment

<!-- gh-comment-id:1771354540 --> @mxyng commented on GitHub (Oct 19, 2023): The 5 minute timeout is not currently user configurable. If you want to keep it alive, you can periodically send an empty `generate` request. Something like this will work: ``` while :; do curl localhost:11434/api/generate -d '{"model":"<model>","system":"","prompt":"","template":""}'; sleep 60; done ``` There are plans on improving the scheduling and session time outs but not at this moment
GiteaMirror commented 2026-05-08 21:52:04 -05:00
Author
Owner
Copy Link

@abdinal1 commented on GitHub (Jan 21, 2024):

@mxyng
is it still the same? I have bíg model and i want to increase the timeout duration form e.g. 5 minutes to 10 minutes

<!-- gh-comment-id:1902721931 --> @abdinal1 commented on GitHub (Jan 21, 2024): @mxyng is it still the same? I have bíg model and i want to increase the timeout duration form e.g. 5 minutes to 10 minutes
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#78078
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?