github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #10348] Issue began after version 0.5.12, Error: POST predict: Post ... wsarecv: An existing connection was forcibly closed by the remote host. #68853

New Issue
Closed
opened 2026-05-04 15:24:48 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-05-04 15:24:48 -05:00
Owner
Copy Link

Originally created by @katmandoo212 on GitHub (Apr 20, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10348

server-2.log

You can see in the server log that the port number is way off, it should be 11434. If I revert back to 0.5.12 the 11434 port is used, but after that version, at least for Windows 10, it attempts to use a different port every execution.

Originally posted by @katmandoo212 in #9986

Originally created by @katmandoo212 on GitHub (Apr 20, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10348 > [server-2.log](https://github.com/user-attachments/files/19824862/server-2.log) > > You can see in the server log that the port number is way off, it should be 11434. If I revert back to 0.5.12 the 11434 port is used, but after that version, at least for Windows 10, it attempts to use a different port every execution. _Originally posted by @katmandoo212 in [#9986](https://github.com/ollama/ollama/issues/9986#issuecomment-2817106369)_
GiteaMirror commented 2026-05-04 15:24:52 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Apr 20, 2025):

time=2025-04-20T05:53:35.864-04:00 level=ERROR source=routes.go:478 msg="embedding generation failed" error="do embedding request: Post \"http://127.0.0.1:62279/embedding\": context canceled"
[GIN] 2025/04/20 - 05:53:35 | 500 |   38.2550735s |       127.0.0.1 | POST     "/api/embed"

The port number is for communicating with the runner, it's not related to the port number of the server (11434). The problem you have is that the context was cancelled and the server returned a 500. The reason for the cancellation is not clear, try adding OLLAMA_DEBUG=1 to the server environment for a more detailed log.

<!-- gh-comment-id:2817114027 --> @rick-github commented on GitHub (Apr 20, 2025): ``` time=2025-04-20T05:53:35.864-04:00 level=ERROR source=routes.go:478 msg="embedding generation failed" error="do embedding request: Post \"http://127.0.0.1:62279/embedding\": context canceled" [GIN] 2025/04/20 - 05:53:35 | 500 | 38.2550735s | 127.0.0.1 | POST "/api/embed" ``` The port number is for communicating with the runner, it's not related to the port number of the server (11434). The problem you have is that the context was cancelled and the server returned a 500. The reason for the cancellation is not clear, try adding `OLLAMA_DEBUG=1` to the server environment for a more detailed log.
GiteaMirror commented 2026-05-04 15:24:52 -05:00
Author
Owner
Copy Link

@katmandoo212 commented on GitHub (Apr 21, 2025):

Here are my logs after installing 0.6.5 and running with OLLAMA_DEBUG=1.

app.log
server.log

<!-- gh-comment-id:2817574533 --> @katmandoo212 commented on GitHub (Apr 21, 2025): Here are my logs after installing 0.6.5 and running with OLLAMA_DEBUG=1. [app.log](https://github.com/user-attachments/files/19829028/app.log) [server.log](https://github.com/user-attachments/files/19829027/server.log)
GiteaMirror commented 2026-05-04 15:24:52 -05:00
Author
Owner
Copy Link

@katmandoo212 commented on GitHub (Apr 21, 2025):

These are the logs from 0.5.12 and running with OLLAMA_DEBUG=1.

app.log
server.log

<!-- gh-comment-id:2817585325 --> @katmandoo212 commented on GitHub (Apr 21, 2025): These are the logs from 0.5.12 and running with OLLAMA_DEBUG=1. [app.log](https://github.com/user-attachments/files/19829105/app.log) [server.log](https://github.com/user-attachments/files/19829106/server.log)
GiteaMirror commented 2026-05-04 15:24:53 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Apr 21, 2025):

D:\home\blt\github\llama.cpp\ggml\src\ggml.c:1729: GGML_ASSERT(tensor->op == GGML_OP_UNARY) failed

#9509
Also:

time=2025-04-20T23:47:29.455-04:00 level=DEBUG source=gpu.go:574 msg="Unable to load cudart library C:\\Users\\User\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\cuda_v11\\cudart64_110.dll: your nvidia driver is too old or missing.  If you have a CUDA GPU please upgrade to run ollama"
time=2025-04-20T23:47:29.464-04:00 level=DEBUG source=gpu.go:574 msg="Unable to load cudart library C:\\Users\\User\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\cuda_v12\\cudart64_12.dll: your nvidia driver is too old or missing.  If you have a CUDA GPU please upgrade to run ollama"

No GPUs are detected, is that expected?

<!-- gh-comment-id:2818999534 --> @rick-github commented on GitHub (Apr 21, 2025): ``` D:\home\blt\github\llama.cpp\ggml\src\ggml.c:1729: GGML_ASSERT(tensor->op == GGML_OP_UNARY) failed ``` #9509 Also: ``` time=2025-04-20T23:47:29.455-04:00 level=DEBUG source=gpu.go:574 msg="Unable to load cudart library C:\\Users\\User\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\cuda_v11\\cudart64_110.dll: your nvidia driver is too old or missing. If you have a CUDA GPU please upgrade to run ollama" time=2025-04-20T23:47:29.464-04:00 level=DEBUG source=gpu.go:574 msg="Unable to load cudart library C:\\Users\\User\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\cuda_v12\\cudart64_12.dll: your nvidia driver is too old or missing. If you have a CUDA GPU please upgrade to run ollama" ``` No GPUs are detected, is that expected?
GiteaMirror commented 2026-05-04 15:24:53 -05:00
Author
Owner
Copy Link

@katmandoo212 commented on GitHub (Apr 21, 2025):

Yes. I am using Intel I5 CPU only.

<!-- gh-comment-id:2819139735 --> @katmandoo212 commented on GitHub (Apr 21, 2025): Yes. I am using Intel I5 CPU only.
GiteaMirror commented 2026-05-04 15:24:54 -05:00
Author
Owner
Copy Link

@katmandoo212 commented on GitHub (Apr 22, 2025):

This issue is resolved on my machine. Here is the breakdown of what resolved it. Thanks so much for your kind help, rick-github!

Suggesting to set OLLAMA_DEBUG and examine the log files did the trick! I saw that the server.log had a line close to the runner termination, that referenced a llama.cpp file that was not in the Ollama installation, but was a llama.cpp source file that I had cloned and downloaded. It stood out because it was on my D drive and not on my C drive.

I had cloned and built llama.cpp from source and had also cloned and built Ollama from source. For some reason, the Ollama build referenced the llama.cpp source code, that I had cloned. I have no clue why. But I resolved the issue by uninstalling my locally built Ollama and moving the llama.cpp source code to another folder (to try to hide it from Ollama). Then I downloaded the Windows Ollama Setup executable for 0.6.5 from the Ollama github. That allows Ollama 0.6.5 to run.

The issue I need to figure out now is this: I am working on an application that I want to be able to use Ollama and llama.cpp in, to compare speeds, and to use the llama.cpp RPC server running on a Raspberry PI cluster. I am only doing this to experiment with, so I expect to run into problems and dead-ends.

<!-- gh-comment-id:2820012732 --> @katmandoo212 commented on GitHub (Apr 22, 2025): This issue is resolved on my machine. Here is the breakdown of what resolved it. Thanks so much for your kind help, rick-github! Suggesting to set OLLAMA_DEBUG and examine the log files did the trick! I saw that the server.log had a line close to the runner termination, that referenced a llama.cpp file that was not in the Ollama installation, but was a llama.cpp source file that I had cloned and downloaded. It stood out because it was on my D drive and not on my C drive. I had cloned and built llama.cpp from source and had also cloned and built Ollama from source. For some reason, the Ollama build referenced the llama.cpp source code, that I had cloned. I have no clue why. But I resolved the issue by uninstalling my locally built Ollama and moving the llama.cpp source code to another folder (to try to hide it from Ollama). Then I downloaded the Windows Ollama Setup executable for 0.6.5 from the Ollama github. That allows Ollama 0.6.5 to run. The issue I need to figure out now is this: I am working on an application that I want to be able to use Ollama and llama.cpp in, to compare speeds, and to use the llama.cpp RPC server running on a Raspberry PI cluster. I am only doing this to experiment with, so I expect to run into problems and dead-ends.
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
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#68853
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?