github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 16:40:08 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #12657] Ollama on Windows spawns a new netsh process every 5 seconds #70458

New Issue
Closed
opened 2026-05-04 21:37:18 -05:00 by GiteaMirror · 8 comments
GiteaMirror commented 2026-05-04 21:37:18 -05:00
Owner
Copy Link

Originally created by @lowleveldesign on GitHub (Oct 16, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12657

Originally assigned to: @BruceMacD on GitHub.

What is the issue?

I observed that on Windows ollama spawn a new netsh process every 5 seconds. Is this really necessary? Is there a way to disable that behavior?

Example wtrace output (you may do the same test with procmon):

 wtrace --handlers process -f 'pname ~ netsh' -f 'pname ~ svchost' -f 'name ~ PROCESS'

...
14:36:59.0928 netsh (26132.-1) Process/Start command line: 'netsh interface show interface'
14:36:59.2548 svchost (8940.-1) Process/Start command line: 'C:\Windows\System32\svchost.exe -k netsvcs -p -s NetSetupSvc'
14:37:04.0905 netsh (14040.-1) Process/Start command line: 'netsh interface show interface'
14:37:04.2606 svchost (13440.-1) Process/Start command line: 'C:\Windows\System32\svchost.exe -k netsvcs -p -s NetSetupSvc'
...

│ ├─ ollama app [11608]
│ │ ├─ ollama [9460]
│ │ │ ├─ conhost [17416]
│ │ │ ├─ ollama [23316]
│ │ │ │ ├─ conhost [24720]
│ │ ├─ netsh [26132] started at 14:36:59.0928  finished at 14:36:59.3321
│ │ │ ├─ conhost [21976] started at 14:36:59.1043  finished at 14:36:59.3359
│ │ ├─ netsh [14040] started at 14:37:04.0905  finished at 14:37:04.3369
│ │ │ ├─ conhost [19532] started at 14:37:04.1021  finished at 14:37:04.3402

Thanks!

Relevant log output

OS

Windows

GPU

No response

CPU

No response

Ollama version

0.12.5

Originally created by @lowleveldesign on GitHub (Oct 16, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12657 Originally assigned to: @BruceMacD on GitHub. ### What is the issue? I observed that on Windows ollama spawn a new netsh process every 5 seconds. Is this really necessary? Is there a way to disable that behavior? Example [wtrace](https://github.com/lowleveldesign/wtrace) output (you may do the same test with procmon): ```sh wtrace --handlers process -f 'pname ~ netsh' -f 'pname ~ svchost' -f 'name ~ PROCESS' ... 14:36:59.0928 netsh (26132.-1) Process/Start command line: 'netsh interface show interface' 14:36:59.2548 svchost (8940.-1) Process/Start command line: 'C:\Windows\System32\svchost.exe -k netsvcs -p -s NetSetupSvc' 14:37:04.0905 netsh (14040.-1) Process/Start command line: 'netsh interface show interface' 14:37:04.2606 svchost (13440.-1) Process/Start command line: 'C:\Windows\System32\svchost.exe -k netsvcs -p -s NetSetupSvc' ... │ ├─ ollama app [11608] │ │ ├─ ollama [9460] │ │ │ ├─ conhost [17416] │ │ │ ├─ ollama [23316] │ │ │ │ ├─ conhost [24720] │ │ ├─ netsh [26132] started at 14:36:59.0928 finished at 14:36:59.3321 │ │ │ ├─ conhost [21976] started at 14:36:59.1043 finished at 14:36:59.3359 │ │ ├─ netsh [14040] started at 14:37:04.0905 finished at 14:37:04.3369 │ │ │ ├─ conhost [19532] started at 14:37:04.1021 finished at 14:37:04.3402 ``` Thanks! ### Relevant log output ```shell ``` ### OS Windows ### GPU _No response_ ### CPU _No response_ ### Ollama version 0.12.5
GiteaMirror added the bug label 2026-05-04 21:37:18 -05:00
GiteaMirror commented 2026-05-04 21:37:25 -05:00
Author
Owner
Copy Link

@Eason023 commented on GitHub (Oct 17, 2025):

@lowleveldesign Hi! I’d like to try to work on this issue.
Plan: add a small throttle/backoff so that netsh isn’t run every 5s on Windows (run only on state-change or with at least a 60s interval), plus an opt-out env var (OLLAMA_DISABLE_NETSH_POLL=1) and a debug log for verification.
If maintainers have a preferred interval or mechanism, I’m happy to adjust.

<!-- gh-comment-id:3413885996 --> @Eason023 commented on GitHub (Oct 17, 2025): @lowleveldesign Hi! I’d like to try to work on this issue. **Plan:** add a small throttle/backoff so that `netsh` isn’t run every 5s on Windows (run only on state-change or with at least a 60s interval), plus an opt-out env var (`OLLAMA_DISABLE_NETSH_POLL=1`) and a debug log for verification. If maintainers have a preferred interval or mechanism, I’m happy to adjust.
GiteaMirror commented 2026-05-04 21:37:29 -05:00
Author
Owner
Copy Link

@levicki commented on GitHub (Oct 17, 2025):

This other bug is related (reports the behavior of svchost.exe launching netsetupsvc):
https://github.com/ollama/ollama/issues/12510

Plan: add a small throttle/backoff so that netsh isn’t run every 5s on Windows

What is the purpose of executing this netsh polling anyway?

If I am not mistaken, Go uses Windows' native way of querying interfaces?

I am not a maintainer and I was unable to locate the offending code in this repo, but to me it looks like a hack, so I don't agree it should be just throttled and/or worked around but actually implemented correctly if possible and removed alltogether if not.

<!-- gh-comment-id:3415144355 --> @levicki commented on GitHub (Oct 17, 2025): This other bug is related (reports the behavior of svchost.exe launching netsetupsvc): https://github.com/ollama/ollama/issues/12510 > Plan: add a small throttle/backoff so that netsh isn’t run every 5s on Windows What is the purpose of executing this netsh polling anyway? If I am not mistaken, Go uses Windows' native way of querying interfaces? I am not a maintainer and I was unable to locate the offending code in this repo, but to me it looks like a hack, so I don't agree it should be just throttled and/or worked around but actually implemented correctly if possible and removed alltogether if not.
GiteaMirror commented 2026-05-04 21:37:35 -05:00
Author
Owner
Copy Link

@BruceMacD commented on GitHub (Oct 21, 2025):

Thanks for bringing this my attention, this code was originally added to try to figure out why models were not responding sometimes in the app. My suspicion was that it was something to do with not being able to download the models back then. I've removed this code for the next release.

<!-- gh-comment-id:3428355915 --> @BruceMacD commented on GitHub (Oct 21, 2025): Thanks for bringing this my attention, this code was originally added to try to figure out why models were not responding sometimes in the app. My suspicion was that it was something to do with not being able to download the models back then. I've removed this code for the next release.
GiteaMirror commented 2026-05-04 21:37:36 -05:00
Author
Owner
Copy Link

@lowleveldesign commented on GitHub (Oct 22, 2025):

Awesome! Thank you, @BruceMacD.

<!-- gh-comment-id:3430730983 --> @lowleveldesign commented on GitHub (Oct 22, 2025): Awesome! Thank you, @BruceMacD.
GiteaMirror commented 2026-05-04 21:37:38 -05:00
Author
Owner
Copy Link

@levicki commented on GitHub (Oct 30, 2025):

Thanks for bringing this my attention, this code was originally added to try to figure out why models were not responding sometimes in the app. My suspicion was that it was something to do with not being able to download the models back then. I've removed this code for the next release.

I am not seeing this referenced in the commit in 0.12.7 release, did you actually remove it?

<!-- gh-comment-id:3468978771 --> @levicki commented on GitHub (Oct 30, 2025): > Thanks for bringing this my attention, this code was originally added to try to figure out why models were not responding sometimes in the app. My suspicion was that it was something to do with not being able to download the models back then. I've removed this code for the next release. I am not seeing this referenced in the commit in 0.12.7 release, did you actually remove it?
GiteaMirror commented 2026-05-04 21:37:39 -05:00
Author
Owner
Copy Link

@BruceMacD commented on GitHub (Oct 30, 2025):

@levicki the app commits aren't in the main repo commit history (yet), so it won't show up there

<!-- gh-comment-id:3469025196 --> @BruceMacD commented on GitHub (Oct 30, 2025): @levicki the app commits aren't in the main repo commit history (yet), so it won't show up there
GiteaMirror commented 2026-05-04 21:37:40 -05:00
Author
Owner
Copy Link

@levicki commented on GitHub (Oct 30, 2025):

@levicki the app commits aren't in the main repo commit history (yet), so it won't show up there

Ok, but was this issue fixed or pushed back to some other release? How do we know except by staring at task manager to see if svchost.exe keeps appearing and disappearing every 5 seconds?

<!-- gh-comment-id:3469265480 --> @levicki commented on GitHub (Oct 30, 2025): > [@levicki](https://github.com/levicki) the app commits aren't in the main repo commit history (yet), so it won't show up there Ok, but was this issue fixed or pushed back to some other release? How do we know except by staring at task manager to see if svchost.exe keeps appearing and disappearing every 5 seconds?
GiteaMirror commented 2026-05-04 21:37:41 -05:00
Author
Owner
Copy Link

@lowleveldesign commented on GitHub (Oct 31, 2025):

I tested the latest release (v0.12.7) and it no longer starts netsh processes. So the reported problem is fixed.

<!-- gh-comment-id:3471739522 --> @lowleveldesign commented on GitHub (Oct 31, 2025): I tested the latest release (v0.12.7) and it no longer starts netsh processes. So the reported problem is fixed.
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#70458
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?