github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-05 23:53:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #5278] Is it possible to start llama server through dynamic dependency library? #3308

New Issue
Closed
opened 2026-04-12 13:52:41 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 13:52:41 -05:00
Owner
Copy Link

Originally created by @leeyiding on GitHub (Jun 25, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/5278

Hi, I'm trying to run Ollama in Nanos Unikernel, Unikernel a single-process operating system.

I found that in 58d95cc9bd, the running mode of llama server changed from loading dynamic dependency libraries to running through subprocess. Therefore, before version 0.1.32, I could run Ollama normally in Nanos, but it could not run normally in later versions.

I would like to ask whether it is possible to continue to provide the method of loading llama server as a dynamic dependency library and avoid using subprocess. Perhaps the two methods can coexist and users can choose by themselves.

Looking forward to your reply, thank you very much.

Originally created by @leeyiding on GitHub (Jun 25, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/5278 Hi, I'm trying to run Ollama in [Nanos Unikernel](https://github.com/nanovms/nanos), Unikernel a single-process operating system. I found that in https://github.com/ollama/ollama/commit/58d95cc9bd446a8209e7388a96c70367cbafd653, the running mode of llama server changed from loading dynamic dependency libraries to running through subprocess. Therefore, before version 0.1.32, I could run Ollama normally in Nanos, but it could not run normally in later versions. I would like to ask whether it is possible to continue to provide the method of loading llama server as a dynamic dependency library and avoid using subprocess. Perhaps the two methods can coexist and users can choose by themselves. Looking forward to your reply, thank you very much.
GiteaMirror added the feature request label 2026-04-12 13:52:41 -05:00
GiteaMirror commented 2026-04-12 13:52:43 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Jul 2, 2024):

Process isolation has helped us improve reliability, and is a cornerstone of the new concurrency support. We are looking at refining how we ingest llama.cpp via PR #5034 however that will still use the subprocess model. It isn't implemented yet, but we've considered a lowest common denominator CPU version linked into the main application (without any AVX* extensions) so it is theoretically possible that could be wired up to skip the subprocess isolation, but inference performance will be slow by comparison the subprocess runners with GPU or vector extension support.

<!-- gh-comment-id:2204730258 --> @dhiltgen commented on GitHub (Jul 2, 2024): Process isolation has helped us improve reliability, and is a cornerstone of the new concurrency support. We are looking at refining how we ingest llama.cpp via PR #5034 however that will still use the subprocess model. It isn't implemented yet, but we've considered a lowest common denominator CPU version linked into the main application (without any AVX* extensions) so it is theoretically possible that could be wired up to skip the subprocess isolation, but inference performance will be slow by comparison the subprocess runners with GPU or vector extension support.
GiteaMirror commented 2026-04-12 13:52:44 -05:00
Author
Owner
Copy Link

@leeyiding commented on GitHub (Jul 3, 2024):

Got it

<!-- gh-comment-id:2204940916 --> @leeyiding commented on GitHub (Jul 3, 2024): Got it
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 feature request
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#3308
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?