Skip to content
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

Jump to bottom

An operation on a socket could not be performed #6869

Closed
ipzzzzz opened this issue Sep 19, 2024 · 7 comments
Closed

An operation on a socket could not be performed #6869

ipzzzzz opened this issue Sep 19, 2024 · 7 comments
Assignees
@dhiltgen
Labels
bug Something isn't working needs more info More information is needed to assist windows

Comments

@ipzzzzz
Copy link

ipzzzzz commented Sep 19, 2024

What is the issue?

The error reported in the terminal: dial tcp [xxx]:443: bind: An operation on a socket could not be performed because the system lacked sufficient buffer space or because a queue was full.

OS

Windows

GPU

Nvidia

CPU

AMD

Ollama version

0.3.10
@ipzzzzz ipzzzzz added the bug Something isn't working label Sep 19, 2024
@ipzzzzz
Copy link
Author

ipzzzzz commented Sep 19, 2024

When I open ollama, the above error occurs, and at this time, other software also cannot connect to the internet. When I turn off ollama, the internet connection of other software returns to normal.

@dhiltgen dhiltgen added needs more info More information is needed to assist windows labels Sep 20, 2024
@dhiltgen
Copy link
Collaborator

Were you trying to pull a model at the time this took place?

Can you share your server logs? https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md

@dhiltgen dhiltgen self-assigned this Sep 20, 2024
@gyassine
Copy link

gyassine commented Sep 23, 2024 •
edited
Loading

I've got the same problem when I download a model (and by the way, it doesn't finish the download). :

other software also cannot connect to the internet

I have to quit the ollama.app : the internet connection of other software also returns to normal.

OS
OSX Sequoia

GPU

CPU
M1

Ollama version
0.3.10

[GIN] 2024/09/23 - 16:19:13 | 200 |    2.522458ms |       127.0.0.1 | GET      "/api/tags"
[GIN] 2024/09/23 - 16:19:15 | 400 |     139.792µs |       127.0.0.1 | POST     "/api/show"
[GIN] 2024/09/23 - 16:20:10 | 400 |     405.291µs |       127.0.0.1 | POST     "/api/show"
[GIN] 2024/09/23 - 16:22:46 | 400 |         320µs |       127.0.0.1 | POST     "/api/show"
time=2024-09-23T16:25:35.759+04:00 level=INFO source=download.go:370 msg="8eeb52dfb3bb part 11 stalled; retrying. If this persists, press ctrl-c to exit, then 'ollama pull' to find a faster connection."
[GIN] 2024/09/23 - 16:25:36 | 400 |     146.208µs |       127.0.0.1 | POST     "/api/show"
time=2024-09-23T16:25:37.248+04:00 level=INFO source=download.go:291 msg="8eeb52dfb3bb part 11 attempt 0 failed: Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/8e/8eeb52dfb3bb9aefdf9d1ef24b3bdbcfbe82238798c4b918278320b6fcef18fe/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240923%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240923T121309Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=0903c13510160a65baa41c27472758a6da15442a4e5fc6b783e7a1f6ea7be748\": dial tcp 172.66.0.236:443: connect: can't assign requested address, retrying in 1s"
time=2024-09-23T16:25:38.270+04:00 level=INFO source=download.go:291 msg="8eeb52dfb3bb part 11 attempt 1 failed: Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/8e/8eeb52dfb3bb9aefdf9d1ef24b3bdbcfbe82238798c4b918278320b6fcef18fe/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240923%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240923T121309Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=0903c13510160a65baa41c27472758a6da15442a4e5fc6b783e7a1f6ea7be748\": dial tcp 172.66.0.236:443: connect: can't assign requested address, retrying in 2s"
time=2024-09-23T16:25:40.295+04:00 level=INFO source=download.go:291 msg="8eeb52dfb3bb part 11 attempt 2 failed: Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/8e/8eeb52dfb3bb9aefdf9d1ef24b3bdbcfbe82238798c4b918278320b6fcef18fe/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240923%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240923T121309Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=0903c13510160a65baa41c27472758a6da15442a4e5fc6b783e7a1f6ea7be748\": dial tcp 172.66.0.236:443: connect: can't assign requested address, retrying in 4s"
time=2024-09-23T16:25:44.326+04:00 level=INFO source=download.go:291 msg="8eeb52dfb3bb part 11 attempt 3 failed: Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/8e/8eeb52dfb3bb9aefdf9d1ef24b3bdbcfbe82238798c4b918278320b6fcef18fe/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240923%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240923T121309Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=0903c13510160a65baa41c27472758a6da15442a4e5fc6b783e7a1f6ea7be748\": dial tcp 172.66.0.236:443: connect: can't assign requested address, retrying in 8s"
time=2024-09-23T16:25:52.353+04:00 level=INFO source=download.go:291 msg="8eeb52dfb3bb part 11 attempt 4 failed: Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/8e/8eeb52dfb3bb9aefdf9d1ef24b3bdbcfbe82238798c4b918278320b6fcef18fe/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240923%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240923T121309Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=0903c13510160a65baa41c27472758a6da15442a4e5fc6b783e7a1f6ea7be748\": dial tcp 172.66.0.236:443: connect: can't assign requested address, retrying in 16s"
time=2024-09-23T16:26:08.374+04:00 level=INFO source=download.go:291 msg="8eeb52dfb3bb part 11 attempt 5 failed: Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/8e/8eeb52dfb3bb9aefdf9d1ef24b3bdbcfbe82238798c4b918278320b6fcef18fe/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240923%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240923T121309Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=0903c13510160a65baa41c27472758a6da15442a4e5fc6b783e7a1f6ea7be748\": dial tcp 172.66.0.236:443: connect: can't assign requested address, retrying in 32s"
[GIN] 2024/09/23 - 16:26:40 | 200 |        13m39s |       127.0.0.1 | POST     "/api/pull"
[GIN] 2024/09/23 - 16:26:40 | 400 |      60.541µs |       127.0.0.1 | POST     "/api/show"

@gyassine
Copy link

Maybe I've got more info on this issue: maybe bandwith... I don't really know how to check...

But at the beginning of download everything is going well, after a few minutes some apps are deconnecting from internet (like teams, etc.) while the download is still going on. After some more minutes, some webpages doesn't work (mostly the one that have not any cache), and the download also stops.

I've got to quit ollama.app to gain access to the net again... otherwise it still prevent me from accessing the net...

@bvwidt
Copy link

bvwidt commented Sep 24, 2024

I have had the same issue. I installed a fresh ollama and I have started it.
However I was unable to download a model and I received the same error messages as above.
Then I shut down ollama and tried hours later. 🌟 TADA... It works now.

When I consult the logs it looks like another server has been started or ollama was stuck in a loop trying to answer endless requests.

See the corresponding log files attached:

@dhiltgen
Copy link
Collaborator

@ipzzzzz it sounds like Ollama might be overwhelming your network. We have a feature request tracking rate limiting for slower network connections with #2006 . In general, Ollama tries to optimize pulling to be as fast as possible given the large size of the models. We're working to improve the behavior on slower networks.

@bvwidt your problem may not be related. If you're seeing multiple instances of Ollama conflicting on windows, you may be hitting #2854 We don't fully understand the root cause on that one but 2 possible scenarios are another Ollama running under another user account on the system, or setting OLLAMA_HOST differently.

@dhiltgen
Copy link
Collaborator

dhiltgen commented Oct 8, 2024

We recently fixed a bug related to retries leaking file descriptors which is likely the root cause of this. This should be fixed in v0.3.13

@dhiltgen dhiltgen closed this as completed Oct 8, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@dhiltgen dhiltgen

Labels
bug Something isn't working needs more info More information is needed to assist windows
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
4 participants
@bvwidt @dhiltgen @gyassine @ipzzzzz