github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-23 14:13:13 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #6679] HTTP_PROXY Not Being Used in Model Requests #4205

New Issue
Closed
opened 2026-04-12 15:08:15 -05:00 by GiteaMirror · 30 comments
GiteaMirror commented 2026-04-12 15:08:15 -05:00
Owner
Copy Link

Originally created by @cmilhaupt on GitHub (Sep 6, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6679

What is the issue?

The HTTP_PROXY and HTTPS_PROXY variables aren't being used when requesting the model manifest file and pulling of the model itself. The symptom that lead me to the extra debugging started when trying to pull a model and getting the following error:

$ ollama pull gemma2:27b
pulling manifest
Error: pull model manifest: invalid character '<' looking for beginning of value

This error occurred for any model I tried to pull. Cloning the software and building according to the developer instructions, I was able to print out the full body of the response which revealed an HTML page (starting with <html>) being returned by my corporate proxy. I have both HTTP_PROXY and HTTPS_PROXY set in /etc/environments and other Go programs are able to see them. What ended up working for me was hard-coding the proxy directly into ollama:

diff --git a/server/download.go b/server/download.go
index 02f7ae88..71ae8329 100644
--- a/server/download.go
+++ b/server/download.go
@@ -330,7 +330,12 @@ func (b *blobDownload) downloadChunk(ctx context.Context, requestURL *url.URL, w
                        return err
                }
                req.Header.Set("Range", fmt.Sprintf("bytes=%d-%d", part.StartsAt(), part.StopsAt()-1))
-               resp, err := http.DefaultClient.Do(req)
+
+               u, err := url.Parse("http://myproxy:port")
+               http_client := &http.Client{
+                       Transport: &http.Transport{Proxy: http.ProxyURL(u)},
+               }
+               resp, err := http_client.Do(req)
                if err != nil {
                        return err
                }
diff --git a/server/images.go b/server/images.go
index b5bf7ad6..bdbd01fd 100644
--- a/server/images.go
+++ b/server/images.go
@@ -1098,9 +1098,14 @@ func makeRequest(ctx context.Context, method string, requestURL *url.URL, header
                req.ContentLength = contentLength
        }

-       resp, err := (&http.Client{
+       u, err := url.Parse("http://myproxy:port")
+       http_client := &http.Client{
+               Transport: &http.Transport{Proxy: http.ProxyURL(u)},
                CheckRedirect: regOpts.CheckRedirect,
-       }).Do(req)
+       }
+
+       resp, err := http_client.Do(req)
+
        if err != nil {
                return nil, err
        }

I'm still new-ish to golang so I'm not sure what the proper fix is here besides knowing it isn't what I've provided. If it helps, I'm using go v1.23.0. Some light googling pointed me to the idea that it's possible to instantiate a default http.Transport struct which should include the proxy information, so I'm sure there are more generic ways to apply this patch. I'm happy to open the MR with a little guidance on what the generic solution would be.

OS

Linux

GPU

Nvidia

CPU

AMD

Ollama version

v0.3.9

Originally created by @cmilhaupt on GitHub (Sep 6, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6679 ### What is the issue? The `HTTP_PROXY` and `HTTPS_PROXY` variables aren't being used when requesting the model manifest file and pulling of the model itself. The symptom that lead me to the extra debugging started when trying to pull a model and getting the following error: ``` $ ollama pull gemma2:27b pulling manifest Error: pull model manifest: invalid character '<' looking for beginning of value ``` This error occurred for any model I tried to pull. Cloning the software and building according to the developer instructions, I was able to print out the full body of the response which revealed an HTML page (starting with `<html>`) being returned by my corporate proxy. I have both `HTTP_PROXY` and `HTTPS_PROXY` set in /etc/environments and other Go programs are able to see them. What ended up working for me was hard-coding the proxy directly into ollama: ``` diff --git a/server/download.go b/server/download.go index 02f7ae88..71ae8329 100644 --- a/server/download.go +++ b/server/download.go @@ -330,7 +330,12 @@ func (b *blobDownload) downloadChunk(ctx context.Context, requestURL *url.URL, w return err } req.Header.Set("Range", fmt.Sprintf("bytes=%d-%d", part.StartsAt(), part.StopsAt()-1)) - resp, err := http.DefaultClient.Do(req) + + u, err := url.Parse("http://myproxy:port") + http_client := &http.Client{ + Transport: &http.Transport{Proxy: http.ProxyURL(u)}, + } + resp, err := http_client.Do(req) if err != nil { return err } diff --git a/server/images.go b/server/images.go index b5bf7ad6..bdbd01fd 100644 --- a/server/images.go +++ b/server/images.go @@ -1098,9 +1098,14 @@ func makeRequest(ctx context.Context, method string, requestURL *url.URL, header req.ContentLength = contentLength } - resp, err := (&http.Client{ + u, err := url.Parse("http://myproxy:port") + http_client := &http.Client{ + Transport: &http.Transport{Proxy: http.ProxyURL(u)}, CheckRedirect: regOpts.CheckRedirect, - }).Do(req) + } + + resp, err := http_client.Do(req) + if err != nil { return nil, err } ``` I'm still new-ish to golang so I'm not sure what the proper fix is here besides knowing it isn't what I've provided. If it helps, I'm using go v1.23.0. Some light googling pointed me to the idea that it's possible to instantiate a default http.Transport struct which should include the proxy information, so I'm sure there are more generic ways to apply this patch. I'm happy to open the MR with a little guidance on what the generic solution would be. ### OS Linux ### GPU Nvidia ### CPU AMD ### Ollama version v0.3.9
GiteaMirror added the bug label 2026-04-12 15:08:15 -05:00
GiteaMirror commented 2026-04-12 15:08:16 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 6, 2024):

Have you set HTTPS_PROXY in the ollama service file (not /etc/environments)? sudo systemctl edit ollama.service and add

[Service]
Environment="HTTPS_PROXY=http://your.proxy.name.here:port"

Quick&dirty local test shows proxying working as expected:

$ curl https://icanhazip.com
80.218.136.70
# proxy server on local lan enabled
$ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com
34.171.115.155
# start olllama server with proxy
$ HTTPS_PROXY=http://proxy-usa:8080 ollama serve 2>/dev/null &
[1] 3700606
$ ollama list
NAME      	ID          	SIZE  	MODIFIED       
# pull a model
$ ollama pull qwen2:0.5b
pulling manifest 
pulling 8de95da68dc4... 100% ▕██████████████████████████████████████████████▏ 352 MB                         
pulling 62fbfd9ed093... 100% ▕██████████████████████████████████████████████▏  182 B                         
pulling c156170b718e... 100% ▕██████████████████████████████████████████████▏  11 KB                         
pulling f02dd72bb242... 100% ▕██████████████████████████████████████████████▏   59 B                         
pulling 2184ab82477b... 100% ▕██████████████████████████████████████████████▏  488 B                         
verifying sha256 digest 
writing manifest 
success 
# proxy turned off, try using it
$ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com
curl: (5) Could not resolve proxy: proxy-usa
# try pulling a model with the proxy down
$ ollama pull qwen2:0.5b
pulling manifest 
Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b": proxyconnect tcp: dial tcp: lookup proxy-usa: no such host
<!-- gh-comment-id:2334372951 --> @rick-github commented on GitHub (Sep 6, 2024): Have you [set](https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-linux) HTTPS_PROXY in the ollama service file (not /etc/environments)? `sudo systemctl edit ollama.service` and add ``` [Service] Environment="HTTPS_PROXY=http://your.proxy.name.here:port" ``` Quick&dirty local test shows proxying working as expected: ```sh $ curl https://icanhazip.com 80.218.136.70 # proxy server on local lan enabled $ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com 34.171.115.155 # start olllama server with proxy $ HTTPS_PROXY=http://proxy-usa:8080 ollama serve 2>/dev/null & [1] 3700606 $ ollama list NAME ID SIZE MODIFIED # pull a model $ ollama pull qwen2:0.5b pulling manifest pulling 8de95da68dc4... 100% ▕██████████████████████████████████████████████▏ 352 MB pulling 62fbfd9ed093... 100% ▕██████████████████████████████████████████████▏ 182 B pulling c156170b718e... 100% ▕██████████████████████████████████████████████▏ 11 KB pulling f02dd72bb242... 100% ▕██████████████████████████████████████████████▏ 59 B pulling 2184ab82477b... 100% ▕██████████████████████████████████████████████▏ 488 B verifying sha256 digest writing manifest success # proxy turned off, try using it $ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com curl: (5) Could not resolve proxy: proxy-usa # try pulling a model with the proxy down $ ollama pull qwen2:0.5b pulling manifest Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b": proxyconnect tcp: dial tcp: lookup proxy-usa: no such host ```
GiteaMirror commented 2026-04-12 15:08:17 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 10, 2024):

I have same issue with newest ollama version with set HTTPS_PROXY variable when doing an ollama serve
It seems that those values are not handled properly.

<!-- gh-comment-id:2340177962 --> @KhazAkar commented on GitHub (Sep 10, 2024): I have same issue with newest ollama version with set `HTTPS_PROXY` variable when doing an `ollama serve` It seems that those values are not handled properly.
GiteaMirror commented 2026-04-12 15:08:17 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 10, 2024):

I performed the same test as above with 0.3.10, with the same results: HTTPS_PROXY works as expected for me. If you could provide server logs, the contents of your ollama service and overrides files, errors returned by ollama pull, and the results of your tests of proxy connectivity, that will aid in debugging.

<!-- gh-comment-id:2340277530 --> @rick-github commented on GitHub (Sep 10, 2024): I performed the same test as above with 0.3.10, with the same results: HTTPS_PROXY works as expected for me. If you could provide [server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues), the contents of your ollama service and overrides files, errors returned by `ollama pull`, and the results of your tests of proxy connectivity, that will aid in debugging.
GiteaMirror commented 2026-04-12 15:08:18 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 10, 2024):

Sure:
ollama pull:

x@y:~$ ollama pull llama3.1:latest
pulling manifest 
Error: pull model manifest: invalid character '<' looking for beginning of value

ollama serve:

x@y:~$ OLLAMA_DEBUG=1 HTTPS_PROXY=http://company-proxy.com:8080 ollama serve
2024/09/10 12:47:26 routes.go:1125: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:true OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/home/z0022112/.ollama/models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
time=2024-09-10T12:47:26.622+02:00 level=INFO source=images.go:753 msg="total blobs: 0"
time=2024-09-10T12:47:26.622+02:00 level=INFO source=images.go:760 msg="total unused blobs removed: 0"
time=2024-09-10T12:47:26.622+02:00 level=INFO source=routes.go:1172 msg="Listening on 127.0.0.1:11434 (version 0.3.10)"
time=2024-09-10T12:47:26.623+02:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2148258639/runners
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libggml.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libllama.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/ollama_llama_server.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libggml.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libllama.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/ollama_llama_server.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libggml.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libllama.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/ollama_llama_server.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libggml.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libllama.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/ollama_llama_server.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libggml.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libllama.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/ollama_llama_server.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libggml.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libllama.so.gz
time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/ollama_llama_server.gz
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cpu/ollama_llama_server
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cpu_avx/ollama_llama_server
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cpu_avx2/ollama_llama_server
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cuda_v11/ollama_llama_server
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cuda_v12/ollama_llama_server
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/rocm_v60102/ollama_llama_server
time=2024-09-10T12:47:37.097+02:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cuda_v12 rocm_v60102 cpu cpu_avx cpu_avx2 cuda_v11]"
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:45 msg="Override detection logic by setting OLLAMA_LLM_LIBRARY"
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=sched.go:105 msg="starting llm scheduler"
time=2024-09-10T12:47:37.097+02:00 level=INFO source=gpu.go:200 msg="looking for compatible GPUs"
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=gpu.go:86 msg="searching for GPU discovery libraries for NVIDIA"
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=gpu.go:468 msg="Searching for GPU library" name=libcuda.so*
time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=gpu.go:491 msg="gpu library search" globs="[/usr/local/lib/ollama/libcuda.so* /home/x/libcuda.so* /usr/local/cuda*/targets/*/lib/libcuda.so* /usr/lib/*-linux-gnu/nvidia/current/libcuda.so* /usr/lib/*-linux-gnu/libcuda.so* /usr/lib/wsl/lib/libcuda.so* /usr/lib/wsl/drivers/*/libcuda.so* /opt/cuda/lib*/libcuda.so* /usr/local/cuda/lib*/libcuda.so* /usr/lib*/libcuda.so* /usr/local/lib*/libcuda.so*]"
time=2024-09-10T12:47:37.105+02:00 level=DEBUG source=gpu.go:525 msg="discovered GPU libraries" paths="[/usr/lib/i386-linux-gnu/libcuda.so.535.161.08 /usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08]"
library /usr/lib/i386-linux-gnu/libcuda.so.535.161.08 load err: /usr/lib/i386-linux-gnu/libcuda.so.535.161.08: wrong ELF class: ELFCLASS32
time=2024-09-10T12:47:37.105+02:00 level=DEBUG source=gpu.go:566 msg="skipping 32bit library" library=/usr/lib/i386-linux-gnu/libcuda.so.535.161.08
CUDA driver version: 12.2
time=2024-09-10T12:47:37.110+02:00 level=DEBUG source=gpu.go:119 msg="detected GPUs" count=1 library=/usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08
[GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA totalMem 2048 mb
[GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA freeMem 1729 mb
[GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] Compute Capability 7.5
time=2024-09-10T12:47:37.321+02:00 level=DEBUG source=amd_linux.go:371 msg="amdgpu driver not detected /sys/module/amdgpu"
releasing cuda driver library
time=2024-09-10T12:47:37.321+02:00 level=INFO source=types.go:107 msg="inference compute" id=GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b library=cuda variant=v12 compute=7.5 driver=12.2 name="GRID RTX6000-2Q" total="2.0 GiB" available="1.7 GiB"
[GIN] 2024/09/10 - 12:48:11 | 200 |      36.037µs |       127.0.0.1 | HEAD     "/"
[GIN] 2024/09/10 - 12:48:12 | 200 |  331.917651ms |       127.0.0.1 | POST     "/api/pull"
<!-- gh-comment-id:2340325465 --> @KhazAkar commented on GitHub (Sep 10, 2024): Sure: ollama pull: ``` x@y:~$ ollama pull llama3.1:latest pulling manifest Error: pull model manifest: invalid character '<' looking for beginning of value ``` ollama serve: ``` x@y:~$ OLLAMA_DEBUG=1 HTTPS_PROXY=http://company-proxy.com:8080 ollama serve 2024/09/10 12:47:26 routes.go:1125: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:true OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/home/z0022112/.ollama/models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" time=2024-09-10T12:47:26.622+02:00 level=INFO source=images.go:753 msg="total blobs: 0" time=2024-09-10T12:47:26.622+02:00 level=INFO source=images.go:760 msg="total unused blobs removed: 0" time=2024-09-10T12:47:26.622+02:00 level=INFO source=routes.go:1172 msg="Listening on 127.0.0.1:11434 (version 0.3.10)" time=2024-09-10T12:47:26.623+02:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2148258639/runners time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libggml.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libllama.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/ollama_llama_server.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libggml.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libllama.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/ollama_llama_server.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libggml.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libllama.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/ollama_llama_server.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libggml.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libllama.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/ollama_llama_server.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libggml.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libllama.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/ollama_llama_server.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libggml.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libllama.so.gz time=2024-09-10T12:47:26.623+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/ollama_llama_server.gz time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cpu/ollama_llama_server time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cpu_avx/ollama_llama_server time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cpu_avx2/ollama_llama_server time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cuda_v11/ollama_llama_server time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/cuda_v12/ollama_llama_server time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama2148258639/runners/rocm_v60102/ollama_llama_server time=2024-09-10T12:47:37.097+02:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cuda_v12 rocm_v60102 cpu cpu_avx cpu_avx2 cuda_v11]" time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=payload.go:45 msg="Override detection logic by setting OLLAMA_LLM_LIBRARY" time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=sched.go:105 msg="starting llm scheduler" time=2024-09-10T12:47:37.097+02:00 level=INFO source=gpu.go:200 msg="looking for compatible GPUs" time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=gpu.go:86 msg="searching for GPU discovery libraries for NVIDIA" time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=gpu.go:468 msg="Searching for GPU library" name=libcuda.so* time=2024-09-10T12:47:37.097+02:00 level=DEBUG source=gpu.go:491 msg="gpu library search" globs="[/usr/local/lib/ollama/libcuda.so* /home/x/libcuda.so* /usr/local/cuda*/targets/*/lib/libcuda.so* /usr/lib/*-linux-gnu/nvidia/current/libcuda.so* /usr/lib/*-linux-gnu/libcuda.so* /usr/lib/wsl/lib/libcuda.so* /usr/lib/wsl/drivers/*/libcuda.so* /opt/cuda/lib*/libcuda.so* /usr/local/cuda/lib*/libcuda.so* /usr/lib*/libcuda.so* /usr/local/lib*/libcuda.so*]" time=2024-09-10T12:47:37.105+02:00 level=DEBUG source=gpu.go:525 msg="discovered GPU libraries" paths="[/usr/lib/i386-linux-gnu/libcuda.so.535.161.08 /usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08]" library /usr/lib/i386-linux-gnu/libcuda.so.535.161.08 load err: /usr/lib/i386-linux-gnu/libcuda.so.535.161.08: wrong ELF class: ELFCLASS32 time=2024-09-10T12:47:37.105+02:00 level=DEBUG source=gpu.go:566 msg="skipping 32bit library" library=/usr/lib/i386-linux-gnu/libcuda.so.535.161.08 CUDA driver version: 12.2 time=2024-09-10T12:47:37.110+02:00 level=DEBUG source=gpu.go:119 msg="detected GPUs" count=1 library=/usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08 [GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA totalMem 2048 mb [GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA freeMem 1729 mb [GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] Compute Capability 7.5 time=2024-09-10T12:47:37.321+02:00 level=DEBUG source=amd_linux.go:371 msg="amdgpu driver not detected /sys/module/amdgpu" releasing cuda driver library time=2024-09-10T12:47:37.321+02:00 level=INFO source=types.go:107 msg="inference compute" id=GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b library=cuda variant=v12 compute=7.5 driver=12.2 name="GRID RTX6000-2Q" total="2.0 GiB" available="1.7 GiB" [GIN] 2024/09/10 - 12:48:11 | 200 | 36.037µs | 127.0.0.1 | HEAD "/" [GIN] 2024/09/10 - 12:48:12 | 200 | 331.917651ms | 127.0.0.1 | POST "/api/pull" ```
GiteaMirror commented 2026-04-12 15:08:19 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 10, 2024):

As of curl https://icanhazip.com - it's blocked in my company as security risk :)
But other, like curl https://www.google.com, works:

curl https://www.google.com
<!doctype html><html itemscope="" itemtype="http://schema.org/WebPage" lang="de"><head><meta content="text/html; [...]

with proxy, also works:

HTTPS_PROXY=http://company-proxy.com:8080 curl https://www.google.com
<!doctype html><html itemscope="" itemtype="http://schema.org/WebPage" lang="de"><head><meta content="text/html; [...]
<!-- gh-comment-id:2340334239 --> @KhazAkar commented on GitHub (Sep 10, 2024): As of `curl https://icanhazip.com` - it's blocked in my company as security risk :) But other, like `curl https://www.google.com`, works: ``` curl https://www.google.com <!doctype html><html itemscope="" itemtype="http://schema.org/WebPage" lang="de"><head><meta content="text/html; [...] ``` with proxy, also works: ``` HTTPS_PROXY=http://company-proxy.com:8080 curl https://www.google.com <!doctype html><html itemscope="" itemtype="http://schema.org/WebPage" lang="de"><head><meta content="text/html; [...] ```
GiteaMirror commented 2026-04-12 15:08:19 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 10, 2024):

What do the following commands return:

curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b

HTTPS_PROXY=http://company-proxy.com:8080 curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b
<!-- gh-comment-id:2340337803 --> @rick-github commented on GitHub (Sep 10, 2024): What do the following commands return: ``` curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b ``` ``` HTTPS_PROXY=http://company-proxy.com:8080 curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b ```
GiteaMirror commented 2026-04-12 15:08:20 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 10, 2024):

x@y:~$ curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b
{"schemaVersion":2,"mediaType":"application/vnd.docker.distribution.manifest.v2+json","config":{"digest":"sha256:2184ab82477bc33a5e08fa209df88f0631a19e686320cce2cfe9e00695b2f0e6","mediaType":"application/vnd.docker.container.image.v1+json","size":488},"layers":[{"digest":"sha256:8de95da68dc485c0889c205384c24642f83ca18d089559c977ffc6a3972a71a8","mediaType":"application/vnd.ollama.image.model","size":352151968},{"digest":"sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef","mediaType":"application/vnd.ollama.image.template","size":182},{"digest":"sha256:c156170b718ec29139d3653d40ed1986fd92fb7e0959b5c71f3c48f62e6636f4","mediaType":"application/vnd.ollama.image.license","size":11344},{"digest":"sha256:f02dd72bb2423204352eabc5637b44d79d17f109fdb510a7c51455892aa2d216","mediaType":"application/vnd.ollama.image.params","size":59}]}
x@y:~$ HTTPS_PROXY=http://company-proxy.com:8080 curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b
{"schemaVersion":2,"mediaType":"application/vnd.docker.distribution.manifest.v2+json","config":{"digest":"sha256:2184ab82477bc33a5e08fa209df88f0631a19e686320cce2cfe9e00695b2f0e6","mediaType":"application/vnd.docker.container.image.v1+json","size":488},"layers":[{"digest":"sha256:8de95da68dc485c0889c205384c24642f83ca18d089559c977ffc6a3972a71a8","mediaType":"application/vnd.ollama.image.model","size":352151968},{"digest":"sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef","mediaType":"application/vnd.ollama.image.template","size":182},{"digest":"sha256:c156170b718ec29139d3653d40ed1986fd92fb7e0959b5c71f3c48f62e6636f4","mediaType":"application/vnd.ollama.image.license","size":11344},{"digest":"sha256:f02dd72bb2423204352eabc5637b44d79d17f109fdb510a7c51455892aa2d216","mediaType":"application/vnd.ollama.image.params","size":59}]}

Additional note: system runs software called netskope.
ollama serve without proxy:

x@y:~$ OLLAMA_DEBUG=1 ollama serve
2024/09/10 12:58:57 routes.go:1125: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:true OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/home/z0022112/.ollama/models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
time=2024-09-10T12:58:57.658+02:00 level=INFO source=images.go:753 msg="total blobs: 0"
time=2024-09-10T12:58:57.659+02:00 level=INFO source=images.go:760 msg="total unused blobs removed: 0"
time=2024-09-10T12:58:57.659+02:00 level=INFO source=routes.go:1172 msg="Listening on 127.0.0.1:11434 (version 0.3.10)"
time=2024-09-10T12:58:57.659+02:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama310528903/runners
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libggml.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libllama.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/ollama_llama_server.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libggml.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libllama.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/ollama_llama_server.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libggml.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libllama.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/ollama_llama_server.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libggml.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libllama.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/ollama_llama_server.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libggml.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libllama.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/ollama_llama_server.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libggml.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libllama.so.gz
time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/ollama_llama_server.gz
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cpu/ollama_llama_server
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cpu_avx/ollama_llama_server
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cpu_avx2/ollama_llama_server
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cuda_v11/ollama_llama_server
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cuda_v12/ollama_llama_server
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/rocm_v60102/ollama_llama_server
time=2024-09-10T12:59:08.173+02:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 cuda_v12 rocm_v60102]"
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:45 msg="Override detection logic by setting OLLAMA_LLM_LIBRARY"
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=sched.go:105 msg="starting llm scheduler"
time=2024-09-10T12:59:08.173+02:00 level=INFO source=gpu.go:200 msg="looking for compatible GPUs"
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=gpu.go:86 msg="searching for GPU discovery libraries for NVIDIA"
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=gpu.go:468 msg="Searching for GPU library" name=libcuda.so*
time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=gpu.go:491 msg="gpu library search" globs="[/usr/local/lib/ollama/libcuda.so* /home/x/libcuda.so* /usr/local/cuda*/targets/*/lib/libcuda.so* /usr/lib/*-linux-gnu/nvidia/current/libcuda.so* /usr/lib/*-linux-gnu/libcuda.so* /usr/lib/wsl/lib/libcuda.so* /usr/lib/wsl/drivers/*/libcuda.so* /opt/cuda/lib*/libcuda.so* /usr/local/cuda/lib*/libcuda.so* /usr/lib*/libcuda.so* /usr/local/lib*/libcuda.so*]"
time=2024-09-10T12:59:08.181+02:00 level=DEBUG source=gpu.go:525 msg="discovered GPU libraries" paths="[/usr/lib/i386-linux-gnu/libcuda.so.535.161.08 /usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08]"
library /usr/lib/i386-linux-gnu/libcuda.so.535.161.08 load err: /usr/lib/i386-linux-gnu/libcuda.so.535.161.08: wrong ELF class: ELFCLASS32
time=2024-09-10T12:59:08.181+02:00 level=DEBUG source=gpu.go:566 msg="skipping 32bit library" library=/usr/lib/i386-linux-gnu/libcuda.so.535.161.08
CUDA driver version: 12.2
time=2024-09-10T12:59:08.186+02:00 level=DEBUG source=gpu.go:119 msg="detected GPUs" count=1 library=/usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08
[GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA totalMem 2048 mb
[GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA freeMem 1729 mb
[GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] Compute Capability 7.5
time=2024-09-10T12:59:08.385+02:00 level=DEBUG source=amd_linux.go:371 msg="amdgpu driver not detected /sys/module/amdgpu"
releasing cuda driver library
time=2024-09-10T12:59:08.385+02:00 level=INFO source=types.go:107 msg="inference compute" id=GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b library=cuda variant=v12 compute=7.5 driver=12.2 name="GRID RTX6000-2Q" total="2.0 GiB" available="1.7 GiB"
[GIN] 2024/09/10 - 12:59:08 | 200 |      40.586µs |       127.0.0.1 | HEAD     "/"
time=2024-09-10T12:59:38.401+02:00 level=INFO source=images.go:1020 msg="request failed: Get \"https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b\": dial tcp 104.21.75.227:443: i/o timeout"
[GIN] 2024/09/10 - 12:59:38 | 200 | 30.015266502s |       127.0.0.1 | POST     "/api/pull"

ollama pull w/o proxy:

x@y:~$ ollama pull qwen2:0.5b
pulling manifest 
Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b": dial tcp 104.21.75.227:443: i/o timeout
<!-- gh-comment-id:2340347074 --> @KhazAkar commented on GitHub (Sep 10, 2024): ``` x@y:~$ curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b {"schemaVersion":2,"mediaType":"application/vnd.docker.distribution.manifest.v2+json","config":{"digest":"sha256:2184ab82477bc33a5e08fa209df88f0631a19e686320cce2cfe9e00695b2f0e6","mediaType":"application/vnd.docker.container.image.v1+json","size":488},"layers":[{"digest":"sha256:8de95da68dc485c0889c205384c24642f83ca18d089559c977ffc6a3972a71a8","mediaType":"application/vnd.ollama.image.model","size":352151968},{"digest":"sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef","mediaType":"application/vnd.ollama.image.template","size":182},{"digest":"sha256:c156170b718ec29139d3653d40ed1986fd92fb7e0959b5c71f3c48f62e6636f4","mediaType":"application/vnd.ollama.image.license","size":11344},{"digest":"sha256:f02dd72bb2423204352eabc5637b44d79d17f109fdb510a7c51455892aa2d216","mediaType":"application/vnd.ollama.image.params","size":59}]} x@y:~$ HTTPS_PROXY=http://company-proxy.com:8080 curl https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b {"schemaVersion":2,"mediaType":"application/vnd.docker.distribution.manifest.v2+json","config":{"digest":"sha256:2184ab82477bc33a5e08fa209df88f0631a19e686320cce2cfe9e00695b2f0e6","mediaType":"application/vnd.docker.container.image.v1+json","size":488},"layers":[{"digest":"sha256:8de95da68dc485c0889c205384c24642f83ca18d089559c977ffc6a3972a71a8","mediaType":"application/vnd.ollama.image.model","size":352151968},{"digest":"sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef","mediaType":"application/vnd.ollama.image.template","size":182},{"digest":"sha256:c156170b718ec29139d3653d40ed1986fd92fb7e0959b5c71f3c48f62e6636f4","mediaType":"application/vnd.ollama.image.license","size":11344},{"digest":"sha256:f02dd72bb2423204352eabc5637b44d79d17f109fdb510a7c51455892aa2d216","mediaType":"application/vnd.ollama.image.params","size":59}]} ``` Additional note: system runs software called netskope. ollama serve without proxy: ``` x@y:~$ OLLAMA_DEBUG=1 ollama serve 2024/09/10 12:58:57 routes.go:1125: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:true OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/home/z0022112/.ollama/models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" time=2024-09-10T12:58:57.658+02:00 level=INFO source=images.go:753 msg="total blobs: 0" time=2024-09-10T12:58:57.659+02:00 level=INFO source=images.go:760 msg="total unused blobs removed: 0" time=2024-09-10T12:58:57.659+02:00 level=INFO source=routes.go:1172 msg="Listening on 127.0.0.1:11434 (version 0.3.10)" time=2024-09-10T12:58:57.659+02:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama310528903/runners time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libggml.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/libllama.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu file=build/linux/x86_64/cpu/bin/ollama_llama_server.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libggml.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/libllama.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx file=build/linux/x86_64/cpu_avx/bin/ollama_llama_server.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libggml.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/libllama.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cpu_avx2 file=build/linux/x86_64/cpu_avx2/bin/ollama_llama_server.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libggml.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/libllama.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v11 file=build/linux/x86_64/cuda_v11/bin/ollama_llama_server.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libggml.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/libllama.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=cuda_v12 file=build/linux/x86_64/cuda_v12/bin/ollama_llama_server.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libggml.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/libllama.so.gz time=2024-09-10T12:58:57.659+02:00 level=DEBUG source=payload.go:182 msg=extracting variant=rocm_v60102 file=build/linux/x86_64/rocm_v60102/bin/ollama_llama_server.gz time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cpu/ollama_llama_server time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cpu_avx/ollama_llama_server time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cpu_avx2/ollama_llama_server time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cuda_v11/ollama_llama_server time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/cuda_v12/ollama_llama_server time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:71 msg="availableServers : found" file=/tmp/ollama310528903/runners/rocm_v60102/ollama_llama_server time=2024-09-10T12:59:08.173+02:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 cuda_v12 rocm_v60102]" time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=payload.go:45 msg="Override detection logic by setting OLLAMA_LLM_LIBRARY" time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=sched.go:105 msg="starting llm scheduler" time=2024-09-10T12:59:08.173+02:00 level=INFO source=gpu.go:200 msg="looking for compatible GPUs" time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=gpu.go:86 msg="searching for GPU discovery libraries for NVIDIA" time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=gpu.go:468 msg="Searching for GPU library" name=libcuda.so* time=2024-09-10T12:59:08.173+02:00 level=DEBUG source=gpu.go:491 msg="gpu library search" globs="[/usr/local/lib/ollama/libcuda.so* /home/x/libcuda.so* /usr/local/cuda*/targets/*/lib/libcuda.so* /usr/lib/*-linux-gnu/nvidia/current/libcuda.so* /usr/lib/*-linux-gnu/libcuda.so* /usr/lib/wsl/lib/libcuda.so* /usr/lib/wsl/drivers/*/libcuda.so* /opt/cuda/lib*/libcuda.so* /usr/local/cuda/lib*/libcuda.so* /usr/lib*/libcuda.so* /usr/local/lib*/libcuda.so*]" time=2024-09-10T12:59:08.181+02:00 level=DEBUG source=gpu.go:525 msg="discovered GPU libraries" paths="[/usr/lib/i386-linux-gnu/libcuda.so.535.161.08 /usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08]" library /usr/lib/i386-linux-gnu/libcuda.so.535.161.08 load err: /usr/lib/i386-linux-gnu/libcuda.so.535.161.08: wrong ELF class: ELFCLASS32 time=2024-09-10T12:59:08.181+02:00 level=DEBUG source=gpu.go:566 msg="skipping 32bit library" library=/usr/lib/i386-linux-gnu/libcuda.so.535.161.08 CUDA driver version: 12.2 time=2024-09-10T12:59:08.186+02:00 level=DEBUG source=gpu.go:119 msg="detected GPUs" count=1 library=/usr/lib/x86_64-linux-gnu/libcuda.so.535.161.08 [GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA totalMem 2048 mb [GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] CUDA freeMem 1729 mb [GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b] Compute Capability 7.5 time=2024-09-10T12:59:08.385+02:00 level=DEBUG source=amd_linux.go:371 msg="amdgpu driver not detected /sys/module/amdgpu" releasing cuda driver library time=2024-09-10T12:59:08.385+02:00 level=INFO source=types.go:107 msg="inference compute" id=GPU-a651a7db-37aa-11b2-ad97-b5f26a68d90b library=cuda variant=v12 compute=7.5 driver=12.2 name="GRID RTX6000-2Q" total="2.0 GiB" available="1.7 GiB" [GIN] 2024/09/10 - 12:59:08 | 200 | 40.586µs | 127.0.0.1 | HEAD "/" time=2024-09-10T12:59:38.401+02:00 level=INFO source=images.go:1020 msg="request failed: Get \"https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b\": dial tcp 104.21.75.227:443: i/o timeout" [GIN] 2024/09/10 - 12:59:38 | 200 | 30.015266502s | 127.0.0.1 | POST "/api/pull" ``` ollama pull w/o proxy: ``` x@y:~$ ollama pull qwen2:0.5b pulling manifest Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b": dial tcp 104.21.75.227:443: i/o timeout ```
GiteaMirror commented 2026-04-12 15:08:21 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 10, 2024):

What's the result of

curl --noproxy "*" https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b

curl -v -D - https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b
<!-- gh-comment-id:2340366968 --> @rick-github commented on GitHub (Sep 10, 2024): What's the result of ``` curl --noproxy "*" https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b ``` ``` curl -v -D - https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b ```
GiteaMirror commented 2026-04-12 15:08:21 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 10, 2024):

I forgot that I had on client side proxy set, hence why previous options worked fine :D

curl -v -D - https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b
*   Trying 172.67.182.229:443...
*   Trying 2606:4700:3036::6815:4be3:443...
* Immediate connect fail for 2606:4700:3036::6815:4be3: Network is unreachable
*   Trying 2606:4700:3034::ac43:b6e5:443...
* Immediate connect fail for 2606:4700:3034::ac43:b6e5: Network is unreachable

with proxy enabled same call:

curl -v -D - https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b
* Uses proxy env variable https_proxy == 'http://company-proxy.com:8080'
*   Trying 163.116.128.80:8080...
* Connected to (nil) (163.116.128.80) port 8080 (#0)
* allocate connect buffer!
* Establish HTTP proxy tunnel to registry.ollama.ai:443
> CONNECT registry.ollama.ai:443 HTTP/1.1
> Host: registry.ollama.ai:443
> User-Agent: curl/7.81.0
> Proxy-Connection: Keep-Alive
> 
< HTTP/1.1 200 Connection Established
HTTP/1.1 200 Connection Established
< 

* Proxy replied 200 to CONNECT request
* CONNECT phase completed!
* ALPN, offering h2
* ALPN, offering http/1.1
*  CAfile: /etc/ssl/certs/ca-certificates.crt
*  CApath: /etc/ssl/certs
* TLSv1.0 (OUT), TLS header, Certificate Status (22):
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
* TLSv1.2 (IN), TLS header, Certificate Status (22):
* TLSv1.3 (IN), TLS handshake, Server hello (2):
* TLSv1.2 (IN), TLS header, Finished (20):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Encrypted Extensions (8):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Certificate (11):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, CERT verify (15):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Finished (20):
* TLSv1.2 (OUT), TLS header, Finished (20):
* TLSv1.3 (OUT), TLS change cipher, Change cipher spec (1):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.3 (OUT), TLS handshake, Finished (20):
* SSL connection using TLSv1.3 / TLS_AES_256_GCM_SHA384
* ALPN, server accepted to use h2
* Server certificate:
*  subject: CN=registry.ollama.ai
*  start date: Aug 11 10:48:11 2024 GMT
*  expire date: Sep 10 10:48:11 2025 GMT
*  subjectAltName: host "registry.ollama.ai" matched cert's "registry.ollama.ai"
*  issuer: C=DE; ST=DE; L=PLACE; O=COMPANY; OU=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX; CN=ca.company.de.goskope.com; emailAddress=certadmin@netskope.com
*  SSL certificate verify ok.
* Using HTTP2, server supports multiplexing
* Connection state changed (HTTP/2 confirmed)
* Copying HTTP/2 data in stream buffer to connection buffer after upgrade: len=0
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* Using Stream ID: 1 (easy handle 0x55bd86fdfeb0)
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
> GET /v2/library/qwen2/manifests/0.5b HTTP/2
> Host: registry.ollama.ai
> user-agent: curl/7.81.0
> accept: */*
> 
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Newsession Ticket (4):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Newsession Ticket (4):
* old SSL session ID is stale, removing
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
< HTTP/2 200 
HTTP/2 200 
< date: Tue, 10 Sep 2024 11:11:25 GMT
date: Tue, 10 Sep 2024 11:11:25 GMT
< content-type: text/plain; charset=utf-8
content-type: text/plain; charset=utf-8
< content-length: 857
content-length: 857
< via: 1.1 google
via: 1.1 google
< cf-cache-status: DYNAMIC
cf-cache-status: DYNAMIC
< report-to: {"endpoints":[{"url":"https:\/\/a.nel.cloudflare.com\/report\/v4?s=688UAOb7Uv12k%2BPdBeHnALQJTBXtNZ%2BKdoRlQHW9PSpyIsQU06QjpTdT0X7Nke2mFAPWUcxwdmQ8D3gMV8B51M4Hs1p9p8zVVYppD9jFwVMqR2LKUJ8ADipN9HFkHoPfi%2FTxwTY%3D"}],"group":"cf-nel","max_age":604800}
report-to: {"endpoints":[{"url":"https:\/\/a.nel.cloudflare.com\/report\/v4?s=688UAOb7Uv12k%2BPdBeHnALQJTBXtNZ%2BKdoRlQHW9PSpyIsQU06QjpTdT0X7Nke2mFAPWUcxwdmQ8D3gMV8B51M4Hs1p9p8zVVYppD9jFwVMqR2LKUJ8ADipN9HFkHoPfi%2FTxwTY%3D"}],"group":"cf-nel","max_age":604800}
< nel: {"success_fraction":0,"report_to":"cf-nel","max_age":604800}
nel: {"success_fraction":0,"report_to":"cf-nel","max_age":604800}
< server: cloudflare
server: cloudflare
< cf-ray: 8c0efa276a3d7174-DUS
cf-ray: 8c0efa276a3d7174-DUS

< 
* TLSv1.2 (IN), TLS header, Supplemental data (23):
{"schemaVersion":2,"mediaType":"application/vnd.docker.distribution.manifest.v2+json","config":{"digest":"sha256:2184ab82477bc33a5e08fa209df88f0631a19e686320cce2cfe9e00695b2f0e6","mediaType":"application/vnd.docker.container.image.v1+json","size":488},"layers":[{"digest":"sha256:8de95da68dc485c0889c205384c24642f83ca18d089559c977ffc6a3972a71a8","mediaType":"application/vnd.ollama.image.model","size":352151968},{"digest":"sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef","mediaType":"application/vnd.ollama.image.template","size":182},{"digest":"sha256:c156170b718ec29139d3653d40ed1986fd92fb7e0959b5c71f3c48f62e6636f4","mediaType":"application/vnd.ollama.image.license","size":11344},{"digest":"sha256:f02dd72bb2423204352eabc5637b44d79d17f109fdb510a7c51455892aa2d216","mediaType":"application/vnd.ollama.image.params","size":59}]}
* Connection #0 to host (nil) left intact
<!-- gh-comment-id:2340379846 --> @KhazAkar commented on GitHub (Sep 10, 2024): I forgot that I had on client side proxy set, hence why previous options worked fine :D ``` curl -v -D - https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b * Trying 172.67.182.229:443... * Trying 2606:4700:3036::6815:4be3:443... * Immediate connect fail for 2606:4700:3036::6815:4be3: Network is unreachable * Trying 2606:4700:3034::ac43:b6e5:443... * Immediate connect fail for 2606:4700:3034::ac43:b6e5: Network is unreachable ``` with proxy enabled same call: ``` curl -v -D - https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b * Uses proxy env variable https_proxy == 'http://company-proxy.com:8080' * Trying 163.116.128.80:8080... * Connected to (nil) (163.116.128.80) port 8080 (#0) * allocate connect buffer! * Establish HTTP proxy tunnel to registry.ollama.ai:443 > CONNECT registry.ollama.ai:443 HTTP/1.1 > Host: registry.ollama.ai:443 > User-Agent: curl/7.81.0 > Proxy-Connection: Keep-Alive > < HTTP/1.1 200 Connection Established HTTP/1.1 200 Connection Established < * Proxy replied 200 to CONNECT request * CONNECT phase completed! * ALPN, offering h2 * ALPN, offering http/1.1 * CAfile: /etc/ssl/certs/ca-certificates.crt * CApath: /etc/ssl/certs * TLSv1.0 (OUT), TLS header, Certificate Status (22): * TLSv1.3 (OUT), TLS handshake, Client hello (1): * TLSv1.2 (IN), TLS header, Certificate Status (22): * TLSv1.3 (IN), TLS handshake, Server hello (2): * TLSv1.2 (IN), TLS header, Finished (20): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Encrypted Extensions (8): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Certificate (11): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, CERT verify (15): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Finished (20): * TLSv1.2 (OUT), TLS header, Finished (20): * TLSv1.3 (OUT), TLS change cipher, Change cipher spec (1): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.3 (OUT), TLS handshake, Finished (20): * SSL connection using TLSv1.3 / TLS_AES_256_GCM_SHA384 * ALPN, server accepted to use h2 * Server certificate: * subject: CN=registry.ollama.ai * start date: Aug 11 10:48:11 2024 GMT * expire date: Sep 10 10:48:11 2025 GMT * subjectAltName: host "registry.ollama.ai" matched cert's "registry.ollama.ai" * issuer: C=DE; ST=DE; L=PLACE; O=COMPANY; OU=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX; CN=ca.company.de.goskope.com; emailAddress=certadmin@netskope.com * SSL certificate verify ok. * Using HTTP2, server supports multiplexing * Connection state changed (HTTP/2 confirmed) * Copying HTTP/2 data in stream buffer to connection buffer after upgrade: len=0 * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * Using Stream ID: 1 (easy handle 0x55bd86fdfeb0) * TLSv1.2 (OUT), TLS header, Supplemental data (23): > GET /v2/library/qwen2/manifests/0.5b HTTP/2 > Host: registry.ollama.ai > user-agent: curl/7.81.0 > accept: */* > * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Newsession Ticket (4): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Newsession Ticket (4): * old SSL session ID is stale, removing * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (IN), TLS header, Supplemental data (23): < HTTP/2 200 HTTP/2 200 < date: Tue, 10 Sep 2024 11:11:25 GMT date: Tue, 10 Sep 2024 11:11:25 GMT < content-type: text/plain; charset=utf-8 content-type: text/plain; charset=utf-8 < content-length: 857 content-length: 857 < via: 1.1 google via: 1.1 google < cf-cache-status: DYNAMIC cf-cache-status: DYNAMIC < report-to: {"endpoints":[{"url":"https:\/\/a.nel.cloudflare.com\/report\/v4?s=688UAOb7Uv12k%2BPdBeHnALQJTBXtNZ%2BKdoRlQHW9PSpyIsQU06QjpTdT0X7Nke2mFAPWUcxwdmQ8D3gMV8B51M4Hs1p9p8zVVYppD9jFwVMqR2LKUJ8ADipN9HFkHoPfi%2FTxwTY%3D"}],"group":"cf-nel","max_age":604800} report-to: {"endpoints":[{"url":"https:\/\/a.nel.cloudflare.com\/report\/v4?s=688UAOb7Uv12k%2BPdBeHnALQJTBXtNZ%2BKdoRlQHW9PSpyIsQU06QjpTdT0X7Nke2mFAPWUcxwdmQ8D3gMV8B51M4Hs1p9p8zVVYppD9jFwVMqR2LKUJ8ADipN9HFkHoPfi%2FTxwTY%3D"}],"group":"cf-nel","max_age":604800} < nel: {"success_fraction":0,"report_to":"cf-nel","max_age":604800} nel: {"success_fraction":0,"report_to":"cf-nel","max_age":604800} < server: cloudflare server: cloudflare < cf-ray: 8c0efa276a3d7174-DUS cf-ray: 8c0efa276a3d7174-DUS < * TLSv1.2 (IN), TLS header, Supplemental data (23): {"schemaVersion":2,"mediaType":"application/vnd.docker.distribution.manifest.v2+json","config":{"digest":"sha256:2184ab82477bc33a5e08fa209df88f0631a19e686320cce2cfe9e00695b2f0e6","mediaType":"application/vnd.docker.container.image.v1+json","size":488},"layers":[{"digest":"sha256:8de95da68dc485c0889c205384c24642f83ca18d089559c977ffc6a3972a71a8","mediaType":"application/vnd.ollama.image.model","size":352151968},{"digest":"sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef","mediaType":"application/vnd.ollama.image.template","size":182},{"digest":"sha256:c156170b718ec29139d3653d40ed1986fd92fb7e0959b5c71f3c48f62e6636f4","mediaType":"application/vnd.ollama.image.license","size":11344},{"digest":"sha256:f02dd72bb2423204352eabc5637b44d79d17f109fdb510a7c51455892aa2d216","mediaType":"application/vnd.ollama.image.params","size":59}]} * Connection #0 to host (nil) left intact ```
GiteaMirror commented 2026-04-12 15:08:22 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 10, 2024):

So it looks like your firewall/proxy product does cert forging and MITM, presumably for deep packet inspection. This should be transparent to the client but perhaps it's not interacting well with the network libraries that ollama uses. I can't find anything on the interwebs about it so it's not a common issue. I also couldn't find any demo version of the netskope software for testing. Colin's fix of hacking the proxy directly in to the Go code means there's a workaround, we just need to figure out what it is.

<!-- gh-comment-id:2340468445 --> @rick-github commented on GitHub (Sep 10, 2024): So it looks like your firewall/proxy product does cert forging and MITM, presumably for deep packet inspection. This should be transparent to the client but perhaps it's not interacting well with the network libraries that ollama uses. I can't find anything on the interwebs about it so it's not a common issue. I also couldn't find any demo version of the netskope software for testing. Colin's fix of hacking the proxy directly in to the Go code means there's a workaround, we just need to figure out what it is.
GiteaMirror commented 2026-04-12 15:08:23 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 10, 2024):

What's the result of env|grep -i proxy?

<!-- gh-comment-id:2341327149 --> @rick-github commented on GitHub (Sep 10, 2024): What's the result of ` env|grep -i proxy`?
GiteaMirror commented 2026-04-12 15:08:23 -05:00
Author
Owner
Copy Link

@WeiguangHan commented on GitHub (Sep 11, 2024):

Have you set HTTPS_PROXY in the ollama service file (not /etc/environments)? sudo systemctl edit ollama.service and add

[Service]
Environment="HTTPS_PROXY=http://your.proxy.name.here:port"

Quick&dirty local test shows proxying working as expected:

$ curl https://icanhazip.com
80.218.136.70
# proxy server on local lan enabled
$ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com
34.171.115.155
# start olllama server with proxy
$ HTTPS_PROXY=http://proxy-usa:8080 ollama serve 2>/dev/null &
[1] 3700606
$ ollama list
NAME      	ID          	SIZE  	MODIFIED       
# pull a model
$ ollama pull qwen2:0.5b
pulling manifest 
pulling 8de95da68dc4... 100% ▕██████████████████████████████████████████████▏ 352 MB                         
pulling 62fbfd9ed093... 100% ▕██████████████████████████████████████████████▏  182 B                         
pulling c156170b718e... 100% ▕██████████████████████████████████████████████▏  11 KB                         
pulling f02dd72bb242... 100% ▕██████████████████████████████████████████████▏   59 B                         
pulling 2184ab82477b... 100% ▕██████████████████████████████████████████████▏  488 B                         
verifying sha256 digest 
writing manifest 
success 
# proxy turned off, try using it
$ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com
curl: (5) Could not resolve proxy: proxy-usa
# try pulling a model with the proxy down
$ ollama pull qwen2:0.5b
pulling manifest 
Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b": proxyconnect tcp: dial tcp: lookup proxy-usa: no such host

It makes sense on my linux machine after I add the env variable. So how can I perform the similar operation on my windows computer? Thanks a lot in advance! @rick-github

<!-- gh-comment-id:2343254671 --> @WeiguangHan commented on GitHub (Sep 11, 2024): > Have you [set](https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-linux) HTTPS_PROXY in the ollama service file (not /etc/environments)? `sudo systemctl edit ollama.service` and add > > ``` > [Service] > Environment="HTTPS_PROXY=http://your.proxy.name.here:port" > ``` > > Quick&dirty local test shows proxying working as expected: > > ```shell > $ curl https://icanhazip.com > 80.218.136.70 > # proxy server on local lan enabled > $ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com > 34.171.115.155 > # start olllama server with proxy > $ HTTPS_PROXY=http://proxy-usa:8080 ollama serve 2>/dev/null & > [1] 3700606 > $ ollama list > NAME ID SIZE MODIFIED > # pull a model > $ ollama pull qwen2:0.5b > pulling manifest > pulling 8de95da68dc4... 100% ▕██████████████████████████████████████████████▏ 352 MB > pulling 62fbfd9ed093... 100% ▕██████████████████████████████████████████████▏ 182 B > pulling c156170b718e... 100% ▕██████████████████████████████████████████████▏ 11 KB > pulling f02dd72bb242... 100% ▕██████████████████████████████████████████████▏ 59 B > pulling 2184ab82477b... 100% ▕██████████████████████████████████████████████▏ 488 B > verifying sha256 digest > writing manifest > success > # proxy turned off, try using it > $ HTTPS_PROXY=http://proxy-usa:8080 curl https://icanhazip.com > curl: (5) Could not resolve proxy: proxy-usa > # try pulling a model with the proxy down > $ ollama pull qwen2:0.5b > pulling manifest > Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/qwen2/manifests/0.5b": proxyconnect tcp: dial tcp: lookup proxy-usa: no such host > ``` It makes sense on my linux machine after I add the env variable. So how can I perform the similar operation on my windows computer? Thanks a lot in advance! @rick-github
GiteaMirror commented 2026-04-12 15:08:24 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 11, 2024):

What's the result of env|grep -i proxy?

@rick-github

After fresh reboot:

http_proxy=http://company-proxy.com:8080
https_proxy=http://company-proxy.com:8080
ftp_proxy=http://company-proxy.com:8080
no_proxy=company.com,old.company.com,login.cloud.company.com,localhost
NO_PROXY=company.com,old.company.com,login.cloud.company.com,localhost

Commented those out on that machine now, did reboot. What to do next?

<!-- gh-comment-id:2343672743 --> @KhazAkar commented on GitHub (Sep 11, 2024): > What's the result of ` env|grep -i proxy`? @rick-github After fresh reboot: ``` http_proxy=http://company-proxy.com:8080 https_proxy=http://company-proxy.com:8080 ftp_proxy=http://company-proxy.com:8080 no_proxy=company.com,old.company.com,login.cloud.company.com,localhost NO_PROXY=company.com,old.company.com,login.cloud.company.com,localhost ``` Commented those out on that machine now, did reboot. What to do next?
GiteaMirror commented 2026-04-12 15:08:25 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Sep 11, 2024):

I haven't been able to reproduce this but I've added some debugging into ollama which should validate if the proxy is set correctly.

Go's default HTTP transport, which clients will automatically use if transport isn't set, already gets proxy configurations (HTTP_PROXY, HTTPS_PROXY, NO_PROXY) from environment variables

<!-- gh-comment-id:2344210162 --> @mxyng commented on GitHub (Sep 11, 2024): I haven't been able to reproduce this but I've added some debugging into ollama which should validate if the proxy is set correctly. Go's [default HTTP transport](https://pkg.go.dev/net/http#DefaultTransport), which clients will automatically use if transport isn't set, already gets proxy configurations (`HTTP_PROXY`, `HTTPS_PROXY`, `NO_PROXY`) from environment variables
GiteaMirror commented 2026-04-12 15:08:25 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 12, 2024):

@WeiguangHan https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-windows

<!-- gh-comment-id:2345036234 --> @rick-github commented on GitHub (Sep 12, 2024): @WeiguangHan https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-windows
GiteaMirror commented 2026-04-12 15:08:26 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 12, 2024):

Wanted to see the value of no_proxy, the go proxy handler has a bunch of rules for skipping using the proxy, but I can't see how these would affect the outcome in this case. Still a mystery as to what's happening.

<!-- gh-comment-id:2345050466 --> @rick-github commented on GitHub (Sep 12, 2024): Wanted to see the value of `no_proxy`, the go [proxy handler](https://cs.opensource.google/go/x/net/+/refs/tags/v0.29.0:http/httpproxy/proxy.go;drc=8c07e20f924fb9dec8d39d2793f72a42c3261a7c;l=169) has a bunch of rules for skipping using the proxy, but I can't see how these would affect the outcome in this case. Still a mystery as to what's happening.
GiteaMirror commented 2026-04-12 15:08:26 -05:00
Author
Owner
Copy Link

@WeiguangHan commented on GitHub (Sep 12, 2024):

@WeiguangHan https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-windows

I set the environment variables according to your suggestion. I still can't using commands ollama pull xxx although I have set the HTTPS_PROXY and https_proxy in my command line. The result is still timeout. I thinks the HTTPS_PROXY are not used in model requests on my windows computer.

<!-- gh-comment-id:2345140537 --> @WeiguangHan commented on GitHub (Sep 12, 2024): > @WeiguangHan https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-windows I set the environment variables according to your suggestion. I still can't using commands ```ollama pull xxx``` although I have set the ```HTTPS_PROXY and https_proxy``` in my command line. The result is still ```timeout```. I thinks the ```HTTPS_PROXY``` are not used in model requests on my windows computer.
GiteaMirror commented 2026-04-12 15:08:27 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 12, 2024):

HTTPS_PROXY, not HTTP_PROXY.

On Thu, Sep 12, 2024, 10:29 WeiguangHan @.***> wrote:

@WeiguangHan https://github.com/WeiguangHan
https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-windows

I set the environment variables according to your suggestion. But I still
can't using commands ollama pull xxx. The result is still timeout. I
thinks the HTTP_PROXY are not used in model requests on my windows
computer.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/6679#issuecomment-2345140537,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ADSBEJTHVQ2RAGLB67P3P5DZWD36ZAVCNFSM6AAAAABNYZFKI2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNBVGE2DANJTG4
.
You are receiving this because you were mentioned.Message ID:
@.***>

<!-- gh-comment-id:2345165828 --> @rick-github commented on GitHub (Sep 12, 2024): HTTPS_PROXY, not HTTP_PROXY. On Thu, Sep 12, 2024, 10:29 WeiguangHan ***@***.***> wrote: > @WeiguangHan <https://github.com/WeiguangHan> > https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-windows > > I set the environment variables according to your suggestion. But I still > can't using commands ollama pull xxx. The result is still timeout. I > thinks the HTTP_PROXY are not used in model requests on my windows > computer. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/6679#issuecomment-2345140537>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ADSBEJTHVQ2RAGLB67P3P5DZWD36ZAVCNFSM6AAAAABNYZFKI2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNBVGE2DANJTG4> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> >
GiteaMirror commented 2026-04-12 15:08:27 -05:00
Author
Owner
Copy Link

@WeiguangHan commented on GitHub (Sep 12, 2024):

Got it. Now it works. Thanks a lot!

<!-- gh-comment-id:2345169500 --> @WeiguangHan commented on GitHub (Sep 12, 2024): Got it. Now it works. Thanks a lot!
GiteaMirror commented 2026-04-12 15:08:28 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 12, 2024):

Wanted to see the value of no_proxy, the go proxy handler has a bunch of rules for skipping using the proxy, but I can't see how these would affect the outcome in this case. Still a mystery as to what's happening.

I've tried cURL on IP returned directly and I get some page that this page doesn't exist. Maybe go proxy code doesn't handle redirects that well? It's a tough nut to crack

<!-- gh-comment-id:2345274513 --> @KhazAkar commented on GitHub (Sep 12, 2024): > Wanted to see the value of `no_proxy`, the go [proxy handler](https://cs.opensource.google/go/x/net/+/refs/tags/v0.29.0:http/httpproxy/proxy.go;drc=8c07e20f924fb9dec8d39d2793f72a42c3261a7c;l=169) has a bunch of rules for skipping using the proxy, but I can't see how these would affect the outcome in this case. Still a mystery as to what's happening. I've tried cURL on IP returned directly and I get some page that this page doesn't exist. Maybe go proxy code doesn't handle redirects that well? It's a tough nut to crack
GiteaMirror commented 2026-04-12 15:08:28 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Sep 12, 2024):

I've tried cURL on IP returned directly and I get some page that this page doesn't exist. Maybe go proxy code doesn't handle redirects that well?

It's unclear what you mean by this. A proxy is not the same thing as a redirect. A request must pass through a forward proxy to reach the Internet. There's no redirect involved here. Something in the proxy is preventing more data from being transferred.

At this point, everything indicates there isn't a bug in Ollama and proxy configurations is working as expected. Since there hasn't any updates from the issue creator @cmilhaupt, the most likely cause of the original issue is misconfiguration:

I have both HTTP_PROXY and HTTPS_PROXY set in /etc/environments and other Go programs are able to see them.

/etc/environment is only applicable to login sessions and ollama server, (likely) running as a systemd service, is unaffected.

FWIW manifests are downloaded directly from the registry and the logs seem to indicate this was successful. Once the manifest is downloaded, it downloads each blob which does contain a redirect to Cloudflare. You should verify this operation:

$ curl -v -L https://registry.ollama.ai/v2/library/qwen2/blobs/sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef
<!-- gh-comment-id:2347424941 --> @mxyng commented on GitHub (Sep 12, 2024): > I've tried cURL on IP returned directly and I get some page that this page doesn't exist. Maybe go proxy code doesn't handle redirects that well? It's unclear what you mean by this. A proxy is not the same thing as a redirect. A request must pass _through_ a forward proxy to reach the Internet. There's no redirect involved here. Something in the proxy is preventing more data from being transferred. At this point, everything indicates there isn't a bug in Ollama and proxy configurations is working as expected. Since there hasn't any updates from the issue creator @cmilhaupt, the most likely cause of the original issue is misconfiguration: > I have both HTTP_PROXY and HTTPS_PROXY set in /etc/environments and other Go programs are able to see them. `/etc/environment` is only applicable to login sessions and ollama server, (likely) running as a systemd service, is unaffected. FWIW manifests are downloaded directly from the registry and the logs seem to indicate this was successful. Once the manifest is downloaded, it downloads each blob which _does_ contain a redirect to Cloudflare. You should verify this operation: ``` $ curl -v -L https://registry.ollama.ai/v2/library/qwen2/blobs/sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef ```
GiteaMirror commented 2026-04-12 15:08:29 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 18, 2024):

With proxy it seems to work, at least this curl call, but I'm not fully sure...

x@y:~$ curl -v -L https://registry.ollama.ai/v2/library/qwen2/blobs/sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef
* Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost'
* Uses proxy env variable https_proxy == 'http://company-proxy.com:8080'
*   Trying 163.116.128.80:8080...
* Connected to (nil) (163.116.128.80) port 8080 (#0)
* allocate connect buffer!
* Establish HTTP proxy tunnel to registry.ollama.ai:443
> CONNECT registry.ollama.ai:443 HTTP/1.1
> Host: registry.ollama.ai:443
> User-Agent: curl/7.81.0
> Proxy-Connection: Keep-Alive
> 
< HTTP/1.1 200 Connection Established
< 
* Proxy replied 200 to CONNECT request
* CONNECT phase completed!
* ALPN, offering h2
* ALPN, offering http/1.1
*  CAfile: /etc/ssl/certs/ca-certificates.crt
*  CApath: /etc/ssl/certs
* TLSv1.0 (OUT), TLS header, Certificate Status (22):
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
* TLSv1.2 (IN), TLS header, Certificate Status (22):
* TLSv1.3 (IN), TLS handshake, Server hello (2):
* TLSv1.2 (IN), TLS header, Finished (20):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Encrypted Extensions (8):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Certificate (11):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, CERT verify (15):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Finished (20):
* TLSv1.2 (OUT), TLS header, Finished (20):
* TLSv1.3 (OUT), TLS change cipher, Change cipher spec (1):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.3 (OUT), TLS handshake, Finished (20):
* SSL connection using TLSv1.3 / TLS_AES_256_GCM_SHA384
* ALPN, server accepted to use h2
* Server certificate:
*  subject: CN=registry.ollama.ai
*  start date: Aug 19 09:38:04 2024 GMT
*  expire date: Sep 18 09:38:04 2025 GMT
*  subjectAltName: host "registry.ollama.ai" matched cert's "registry.ollama.ai"
*  issuer: C=DE; ST=DE; L=X; O=COMPANY; OU=XXXXXXXxXXXXXXXXXXXXXXXXXXXXXXXX; CN=ca.XXX.XXX.goskope.com; emailAddress=certadmin@netskope.com
*  SSL certificate verify ok.
* Using HTTP2, server supports multiplexing
* Connection state changed (HTTP/2 confirmed)
* Copying HTTP/2 data in stream buffer to connection buffer after upgrade: len=0
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* Using Stream ID: 1 (easy handle 0x5568251f7eb0)
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
> GET /v2/library/qwen2/blobs/sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef HTTP/2
> Host: registry.ollama.ai
> user-agent: curl/7.81.0
> accept: */*
> 
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Newsession Ticket (4):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Newsession Ticket (4):
* old SSL session ID is stale, removing
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
< HTTP/2 307 
< date: Wed, 18 Sep 2024 09:38:04 GMT
< content-type: text/html; charset=utf-8
< location: https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/62/62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240918%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240918T093804Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=44fc0f5f281a376e1504c87d111f6bd977529345a72cb4292d70ec8215f99a3e
< via: 1.1 google
< cf-cache-status: DYNAMIC
< report-to: {"endpoints":[{"url":"https:\/\/a.nel.cloudflare.com\/report\/v4?s=bya%2Fun3UWMvJfKIPj7HDjQ48phwyLL2nfzD19aycptBde1CPr4tmtJkDD%2BTXhaDwDZFxrkX6%2FmZWkxieXenIdFtqUZSizGxZRrXDwEP7qlnwbNcsBmswjDUgvhEjNGOGlfg%2Bw%2B4%3D"}],"group":"cf-nel","max_age":604800}
< nel: {"success_fraction":0,"report_to":"cf-nel","max_age":604800}
< speculation-rules: "/cdn-cgi/speculation"
< server: cloudflare
< cf-ray: 8c505c69cb8ccb1e-DUS
< 
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* Ignoring the response-body
* Connection #0 to host (nil) left intact
* Issue another request to this URL: 'https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/62/62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240918%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240918T093804Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=44fc0f5f281a376e1504c87d111f6bd977529345a72cb4292d70ec8215f99a3e'
* Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost'
* Uses proxy env variable https_proxy == 'http://company-proxy.com:8080'
* Hostname company-proxy.com was found in DNS cache
*   Trying 163.116.128.80:8080...
* Connected to (nil) (163.116.128.80) port 8080 (#1)
* allocate connect buffer!
* Establish HTTP proxy tunnel to dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com:443
> CONNECT dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com:443 HTTP/1.1
> Host: dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com:443
> User-Agent: curl/7.81.0
> Proxy-Connection: Keep-Alive
> 
< HTTP/1.1 200 Connection Established
< 
* Proxy replied 200 to CONNECT request
* CONNECT phase completed!
* ALPN, offering h2
* ALPN, offering http/1.1
*  CAfile: /etc/ssl/certs/ca-certificates.crt
*  CApath: /etc/ssl/certs
* TLSv1.0 (OUT), TLS header, Certificate Status (22):
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
* TLSv1.2 (IN), TLS header, Certificate Status (22):
* TLSv1.3 (IN), TLS handshake, Server hello (2):
* TLSv1.2 (IN), TLS header, Finished (20):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Encrypted Extensions (8):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Certificate (11):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, CERT verify (15):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Finished (20):
* TLSv1.2 (OUT), TLS header, Finished (20):
* TLSv1.3 (OUT), TLS change cipher, Change cipher spec (1):
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
* TLSv1.3 (OUT), TLS handshake, Finished (20):
* SSL connection using TLSv1.3 / TLS_AES_256_GCM_SHA384
* ALPN, server did not agree to a protocol
* Server certificate:
*  subject: CN=dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com
*  start date: Aug 19 09:38:04 2024 GMT
*  expire date: Sep 18 09:38:04 2025 GMT
*  subjectAltName: host "dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com" matched cert's "dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com"
*  issuer: C=DE; ST=DE; L=X; O=COMPANY; OU=XXXXXXXxXXXXXXXXXXXXXXXXXXXXXXXX; CN=ca.XXX.XXX.goskope.com; emailAddress=certadmin@netskope.com
*  SSL certificate verify ok.
* TLSv1.2 (OUT), TLS header, Supplemental data (23):
> GET /ollama/docker/registry/v2/blobs/sha256/62/62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240918%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240918T093804Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=44fc0f5f281a376e1504c87d111f6bd977529345a72cb4292d70ec8215f99a3e HTTP/1.1
> Host: dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com
> User-Agent: curl/7.81.0
> Accept: */*
> 
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Newsession Ticket (4):
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* TLSv1.3 (IN), TLS handshake, Newsession Ticket (4):
* old SSL session ID is stale, removing
* TLSv1.2 (IN), TLS header, Supplemental data (23):
* Mark bundle as not supporting multiuse
< HTTP/1.1 200 OK
< Date: Wed, 18 Sep 2024 09:38:05 GMT
< Content-Type: application/octet-stream
< Connection: keep-alive
< Accept-Ranges: bytes
< ETag: "b6a10447e4503d042d486fca41f65161"
< Last-Modified: Sat, 17 Feb 2024 22:53:05 GMT
< Server: cloudflare
< CF-RAY: 8c505c6c4da0cb2f-DUS
< Content-Length: 182
< 
* TLSv1.2 (IN), TLS header, Supplemental data (23):
{{ if .System }}<|im_start|>system
{{ .System }}<|im_end|>
{{ end }}{{ if .Prompt }}<|im_start|>user
{{ .Prompt }}<|im_end|>
{{ end }}<|im_start|>assistant
{{ .Response }}<|im_end|>
* Connection #1 to host (nil) left intact
<!-- gh-comment-id:2357994773 --> @KhazAkar commented on GitHub (Sep 18, 2024): With proxy it seems to work, at least this curl call, but I'm not fully sure... ``` x@y:~$ curl -v -L https://registry.ollama.ai/v2/library/qwen2/blobs/sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef * Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost' * Uses proxy env variable https_proxy == 'http://company-proxy.com:8080' * Trying 163.116.128.80:8080... * Connected to (nil) (163.116.128.80) port 8080 (#0) * allocate connect buffer! * Establish HTTP proxy tunnel to registry.ollama.ai:443 > CONNECT registry.ollama.ai:443 HTTP/1.1 > Host: registry.ollama.ai:443 > User-Agent: curl/7.81.0 > Proxy-Connection: Keep-Alive > < HTTP/1.1 200 Connection Established < * Proxy replied 200 to CONNECT request * CONNECT phase completed! * ALPN, offering h2 * ALPN, offering http/1.1 * CAfile: /etc/ssl/certs/ca-certificates.crt * CApath: /etc/ssl/certs * TLSv1.0 (OUT), TLS header, Certificate Status (22): * TLSv1.3 (OUT), TLS handshake, Client hello (1): * TLSv1.2 (IN), TLS header, Certificate Status (22): * TLSv1.3 (IN), TLS handshake, Server hello (2): * TLSv1.2 (IN), TLS header, Finished (20): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Encrypted Extensions (8): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Certificate (11): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, CERT verify (15): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Finished (20): * TLSv1.2 (OUT), TLS header, Finished (20): * TLSv1.3 (OUT), TLS change cipher, Change cipher spec (1): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.3 (OUT), TLS handshake, Finished (20): * SSL connection using TLSv1.3 / TLS_AES_256_GCM_SHA384 * ALPN, server accepted to use h2 * Server certificate: * subject: CN=registry.ollama.ai * start date: Aug 19 09:38:04 2024 GMT * expire date: Sep 18 09:38:04 2025 GMT * subjectAltName: host "registry.ollama.ai" matched cert's "registry.ollama.ai" * issuer: C=DE; ST=DE; L=X; O=COMPANY; OU=XXXXXXXxXXXXXXXXXXXXXXXXXXXXXXXX; CN=ca.XXX.XXX.goskope.com; emailAddress=certadmin@netskope.com * SSL certificate verify ok. * Using HTTP2, server supports multiplexing * Connection state changed (HTTP/2 confirmed) * Copying HTTP/2 data in stream buffer to connection buffer after upgrade: len=0 * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * Using Stream ID: 1 (easy handle 0x5568251f7eb0) * TLSv1.2 (OUT), TLS header, Supplemental data (23): > GET /v2/library/qwen2/blobs/sha256:62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef HTTP/2 > Host: registry.ollama.ai > user-agent: curl/7.81.0 > accept: */* > * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Newsession Ticket (4): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Newsession Ticket (4): * old SSL session ID is stale, removing * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.2 (IN), TLS header, Supplemental data (23): < HTTP/2 307 < date: Wed, 18 Sep 2024 09:38:04 GMT < content-type: text/html; charset=utf-8 < location: https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/62/62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240918%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240918T093804Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=44fc0f5f281a376e1504c87d111f6bd977529345a72cb4292d70ec8215f99a3e < via: 1.1 google < cf-cache-status: DYNAMIC < report-to: {"endpoints":[{"url":"https:\/\/a.nel.cloudflare.com\/report\/v4?s=bya%2Fun3UWMvJfKIPj7HDjQ48phwyLL2nfzD19aycptBde1CPr4tmtJkDD%2BTXhaDwDZFxrkX6%2FmZWkxieXenIdFtqUZSizGxZRrXDwEP7qlnwbNcsBmswjDUgvhEjNGOGlfg%2Bw%2B4%3D"}],"group":"cf-nel","max_age":604800} < nel: {"success_fraction":0,"report_to":"cf-nel","max_age":604800} < speculation-rules: "/cdn-cgi/speculation" < server: cloudflare < cf-ray: 8c505c69cb8ccb1e-DUS < * TLSv1.2 (IN), TLS header, Supplemental data (23): * Ignoring the response-body * Connection #0 to host (nil) left intact * Issue another request to this URL: 'https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/62/62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240918%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240918T093804Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=44fc0f5f281a376e1504c87d111f6bd977529345a72cb4292d70ec8215f99a3e' * Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost' * Uses proxy env variable https_proxy == 'http://company-proxy.com:8080' * Hostname company-proxy.com was found in DNS cache * Trying 163.116.128.80:8080... * Connected to (nil) (163.116.128.80) port 8080 (#1) * allocate connect buffer! * Establish HTTP proxy tunnel to dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com:443 > CONNECT dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com:443 HTTP/1.1 > Host: dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com:443 > User-Agent: curl/7.81.0 > Proxy-Connection: Keep-Alive > < HTTP/1.1 200 Connection Established < * Proxy replied 200 to CONNECT request * CONNECT phase completed! * ALPN, offering h2 * ALPN, offering http/1.1 * CAfile: /etc/ssl/certs/ca-certificates.crt * CApath: /etc/ssl/certs * TLSv1.0 (OUT), TLS header, Certificate Status (22): * TLSv1.3 (OUT), TLS handshake, Client hello (1): * TLSv1.2 (IN), TLS header, Certificate Status (22): * TLSv1.3 (IN), TLS handshake, Server hello (2): * TLSv1.2 (IN), TLS header, Finished (20): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Encrypted Extensions (8): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Certificate (11): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, CERT verify (15): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Finished (20): * TLSv1.2 (OUT), TLS header, Finished (20): * TLSv1.3 (OUT), TLS change cipher, Change cipher spec (1): * TLSv1.2 (OUT), TLS header, Supplemental data (23): * TLSv1.3 (OUT), TLS handshake, Finished (20): * SSL connection using TLSv1.3 / TLS_AES_256_GCM_SHA384 * ALPN, server did not agree to a protocol * Server certificate: * subject: CN=dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com * start date: Aug 19 09:38:04 2024 GMT * expire date: Sep 18 09:38:04 2025 GMT * subjectAltName: host "dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com" matched cert's "dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com" * issuer: C=DE; ST=DE; L=X; O=COMPANY; OU=XXXXXXXxXXXXXXXXXXXXXXXXXXXXXXXX; CN=ca.XXX.XXX.goskope.com; emailAddress=certadmin@netskope.com * SSL certificate verify ok. * TLSv1.2 (OUT), TLS header, Supplemental data (23): > GET /ollama/docker/registry/v2/blobs/sha256/62/62fbfd9ed093d6e5ac83190c86eec5369317919f4b149598d2dbb38900e9faef/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=66040c77ac1b787c3af820529859349a%2F20240918%2Fauto%2Fs3%2Faws4_request&X-Amz-Date=20240918T093804Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=44fc0f5f281a376e1504c87d111f6bd977529345a72cb4292d70ec8215f99a3e HTTP/1.1 > Host: dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com > User-Agent: curl/7.81.0 > Accept: */* > * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Newsession Ticket (4): * TLSv1.2 (IN), TLS header, Supplemental data (23): * TLSv1.3 (IN), TLS handshake, Newsession Ticket (4): * old SSL session ID is stale, removing * TLSv1.2 (IN), TLS header, Supplemental data (23): * Mark bundle as not supporting multiuse < HTTP/1.1 200 OK < Date: Wed, 18 Sep 2024 09:38:05 GMT < Content-Type: application/octet-stream < Connection: keep-alive < Accept-Ranges: bytes < ETag: "b6a10447e4503d042d486fca41f65161" < Last-Modified: Sat, 17 Feb 2024 22:53:05 GMT < Server: cloudflare < CF-RAY: 8c505c6c4da0cb2f-DUS < Content-Length: 182 < * TLSv1.2 (IN), TLS header, Supplemental data (23): {{ if .System }}<|im_start|>system {{ .System }}<|im_end|> {{ end }}{{ if .Prompt }}<|im_start|>user {{ .Prompt }}<|im_end|> {{ end }}<|im_start|>assistant {{ .Response }}<|im_end|> * Connection #1 to host (nil) left intact ```
GiteaMirror commented 2026-04-12 15:08:29 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 18, 2024):

Its unclear what you mean by this. A proxy is not the same thing as a redirect. A request must pass through a forward proxy to reach the Internet. There's no redirect involved here. Something in the proxy is preventing more data from being transferred.

Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/llama3.1/manifests/8b": dial tcp 104.21.75.227:443: i/o timeout
If I do curl to this IP address pointed by ollama when it fails, with proxy enabled, this is what I get:

x@y:~$ curl -v -L 104.21.75.227:443
* Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost'
* Uses proxy env variable https_proxy == 'http://company-proxy.com:8080'
*   Trying 163.116.128.80:8080...
* Connected to (nil) (163.116.128.80) port 8080 (#0)
> GET http://104.21.75.227:443/ HTTP/1.1
> Host: 104.21.75.227:443
> User-Agent: curl/7.81.0
> Accept: */*
> Proxy-Connection: Keep-Alive
> 
* Mark bundle as not supporting multiuse
< HTTP/1.1 502 Bad Gateway
< Content-Length: 1219
< Connection: close
< 
<!DOCTYPE html>
<html><head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">

<title>Bad Gateway</title>
<style type="text/css">
body {
    font-family: Lato, Arial, sans-serif;
    margin: 0;
    padding: 0;
}
#notification {
    padding: 32px;
}
h1 {
    font-size: 20px;
    line-height: 24px;
    font-weight: normal;
    color: black;
    margin: 0 0 16px 0;
}
h2 {
    font-size: 14px;
    line-height: 20px;
    font-weight: normal;
    color: black;
    margin: 0 0 16px 0;
}
p {
    font-size: 12px;
    line-height: 16px;
    color: #646464;
    margin: 0 0 16px 0;
}
</style>
</head>

<body>
<div id="notification">
    <h1>HsFailure (error:14094410:SSL routines:ssl3_read_bytes:sslv3 alert handshake failure)</h1>

    <h2>The destination 104.21.75.227 is not reachable.<br />Contact your IT administrator with the following error:</h2>

    <p>mode='CLIENT', lname='gre-gateway-proxy', rip='104.21.75.227', rp='443', oip='', op='0', lip='163.116.173.54', lp='60994', sni='104.21.75.227', fd='33346', SSL handshake error '-1' sslerr='1' sslerrdesc='error:14094410:SSL routines:ssl3_read_bytes:sslv3 alert handshake failure' sslerrfunc='148' sslerrreason='1040'</p>
</div>
</body></html>
* Closing connection 0

If I try https://ip:

x@y:~$ curl -v -L https://104.21.75.227
* Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost'
* Uses proxy env variable https_proxy == 'http://company-proxy.com:8080'
*   Trying 163.116.128.80:8080...
* Connected to (nil) (163.116.128.80) port 8080 (#0)
* allocate connect buffer!
* Establish HTTP proxy tunnel to 104.21.75.227:443
> CONNECT 104.21.75.227:443 HTTP/1.1
> Host: 104.21.75.227:443
> User-Agent: curl/7.81.0
> Proxy-Connection: Keep-Alive
> 
< HTTP/1.1 200 Connection Established
< 
* Proxy replied 200 to CONNECT request
* CONNECT phase completed!
* ALPN, offering h2
* ALPN, offering http/1.1
*  CAfile: /etc/ssl/certs/ca-certificates.crt
*  CApath: /etc/ssl/certs
* TLSv1.0 (OUT), TLS header, Certificate Status (22):
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
* TLSv1.0 (IN), TLS header, Unknown (21):
* TLSv1.3 (IN), TLS alert, handshake failure (552):
* error:0A000410:SSL routines::sslv3 alert handshake failure
* Closing connection 0
curl: (35) error:0A000410:SSL routines::sslv3 alert handshake failure
<!-- gh-comment-id:2358004050 --> @KhazAkar commented on GitHub (Sep 18, 2024): > Its unclear what you mean by this. A proxy is not the same thing as a redirect. A request must pass _through_ a forward proxy to reach the Internet. There's no redirect involved here. Something in the proxy is preventing more data from being transferred. `Error: pull model manifest: Get "https://registry.ollama.ai/v2/library/llama3.1/manifests/8b": dial tcp 104.21.75.227:443: i/o timeout` If I do curl to this IP address pointed by ollama when it fails, with proxy enabled, this is what I get: ``` x@y:~$ curl -v -L 104.21.75.227:443 * Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost' * Uses proxy env variable https_proxy == 'http://company-proxy.com:8080' * Trying 163.116.128.80:8080... * Connected to (nil) (163.116.128.80) port 8080 (#0) > GET http://104.21.75.227:443/ HTTP/1.1 > Host: 104.21.75.227:443 > User-Agent: curl/7.81.0 > Accept: */* > Proxy-Connection: Keep-Alive > * Mark bundle as not supporting multiuse < HTTP/1.1 502 Bad Gateway < Content-Length: 1219 < Connection: close < <!DOCTYPE html> <html><head> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> <title>Bad Gateway</title> <style type="text/css"> body { font-family: Lato, Arial, sans-serif; margin: 0; padding: 0; } #notification { padding: 32px; } h1 { font-size: 20px; line-height: 24px; font-weight: normal; color: black; margin: 0 0 16px 0; } h2 { font-size: 14px; line-height: 20px; font-weight: normal; color: black; margin: 0 0 16px 0; } p { font-size: 12px; line-height: 16px; color: #646464; margin: 0 0 16px 0; } </style> </head> <body> <div id="notification"> <h1>HsFailure (error:14094410:SSL routines:ssl3_read_bytes:sslv3 alert handshake failure)</h1> <h2>The destination 104.21.75.227 is not reachable.<br />Contact your IT administrator with the following error:</h2> <p>mode='CLIENT', lname='gre-gateway-proxy', rip='104.21.75.227', rp='443', oip='', op='0', lip='163.116.173.54', lp='60994', sni='104.21.75.227', fd='33346', SSL handshake error '-1' sslerr='1' sslerrdesc='error:14094410:SSL routines:ssl3_read_bytes:sslv3 alert handshake failure' sslerrfunc='148' sslerrreason='1040'</p> </div> </body></html> * Closing connection 0 ``` If I try `https://ip`: ``` x@y:~$ curl -v -L https://104.21.75.227 * Uses proxy env variable no_proxy == 'company.com,old.company.com,login.cloud.company.com,other.company.com,another.company.io,localhost' * Uses proxy env variable https_proxy == 'http://company-proxy.com:8080' * Trying 163.116.128.80:8080... * Connected to (nil) (163.116.128.80) port 8080 (#0) * allocate connect buffer! * Establish HTTP proxy tunnel to 104.21.75.227:443 > CONNECT 104.21.75.227:443 HTTP/1.1 > Host: 104.21.75.227:443 > User-Agent: curl/7.81.0 > Proxy-Connection: Keep-Alive > < HTTP/1.1 200 Connection Established < * Proxy replied 200 to CONNECT request * CONNECT phase completed! * ALPN, offering h2 * ALPN, offering http/1.1 * CAfile: /etc/ssl/certs/ca-certificates.crt * CApath: /etc/ssl/certs * TLSv1.0 (OUT), TLS header, Certificate Status (22): * TLSv1.3 (OUT), TLS handshake, Client hello (1): * TLSv1.0 (IN), TLS header, Unknown (21): * TLSv1.3 (IN), TLS alert, handshake failure (552): * error:0A000410:SSL routines::sslv3 alert handshake failure * Closing connection 0 curl: (35) error:0A000410:SSL routines::sslv3 alert handshake failure ```
GiteaMirror commented 2026-04-12 15:08:30 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 18, 2024):

That's the problem, fetching the manifest with curl via proxy works, but apparently the same config doesn't work for your ollama server. Proxying works for others, and looking at the source code hasn't turned up a reason for why proxying in your particular config (company proxy using netskope) doesn't.

Trying just the IP address doesn't work because ollama.com is using a Cloudfare load balancer and it needs a Host header to determine what website to connect to.

As Michael pointed out the most likely cause of the initial post was mis-configuration. In your case this is likely the same, but so far we haven't been able to pinpoint what the actual misconfig is.

Based on your last post, your HTTPS_PROXY is un-encrypted, so it might be useful to see that portion of the handshake that's in cleartext.

sudo apt install tcpflow
sudo tcpflow -c -i any port 8080

Now in a new terminal,, run ollama pull llama3.1:latest and in the tcpflow terminal you should see ollama try to connect to the ollama website via the proxy on port 8080. I don't know if there will be anything useful there, but at this point we've exhausted the debugging possibilities without access to an environment the same as yours.

<!-- gh-comment-id:2358207225 --> @rick-github commented on GitHub (Sep 18, 2024): That's the problem, fetching the manifest with curl via proxy works, but apparently the same config [doesn't work](https://github.com/ollama/ollama/issues/6679#issuecomment-2340325465) for your ollama server. Proxying works for others, and looking at the source code hasn't turned up a reason for why proxying in your particular config (company proxy using netskope) doesn't. Trying just the IP address doesn't work because ollama.com is using a Cloudfare load balancer and it needs a `Host` header to determine what website to connect to. As Michael pointed out the most likely cause of the initial post was mis-configuration. In your case this is likely the same, but so far we haven't been able to pinpoint what the actual misconfig is. Based on your last post, your HTTPS_PROXY is un-encrypted, so it might be useful to see that portion of the handshake that's in cleartext. ``` sudo apt install tcpflow sudo tcpflow -c -i any port 8080 ``` Now in a new terminal,, run `ollama pull llama3.1:latest` and in the `tcpflow` terminal you should see ollama try to connect to the ollama website via the proxy on port 8080. I don't know if there will be anything useful there, but at this point we've exhausted the debugging possibilities without access to an environment the same as yours.
GiteaMirror commented 2026-04-12 15:08:30 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 18, 2024):

You know what? I think I got this. Apparently I need to open web browser and click some stupid button in order to make it work with proxy, periodically. I've run tcpflow command and got output. Should I send it or what? :)

'you must log in to this network before you can access the internet'... yeah.

So it also acts as captive portal. Is there any way to accept it from CLI?

<!-- gh-comment-id:2358231062 --> @KhazAkar commented on GitHub (Sep 18, 2024): You know what? I think I got this. Apparently I need to open web browser and click some stupid button in order to make it work with proxy, periodically. I've run tcpflow command and got output. Should I send it or what? :) 'you must log in to this network before you can access the internet'... yeah. So it also acts as captive portal. Is there any way to accept it from CLI?
GiteaMirror commented 2026-04-12 15:08:30 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 18, 2024):

Ah, that would explain it. AFAIK there's no way for a client like ollama to negotiate a captive portal, you probably need to talk to your IT folk about whitelisting or some other solution.

<!-- gh-comment-id:2358252387 --> @rick-github commented on GitHub (Sep 18, 2024): Ah, that would explain it. AFAIK there's no way for a client like ollama to negotiate a captive portal, you probably need to talk to your IT folk about whitelisting or some other solution.
GiteaMirror commented 2026-04-12 15:08:31 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 18, 2024):

I know that ollama won't be able to do so, it's understandable :)
I've asked in internal communication if somebody have cURL way of negotiating captive portal :)
Thanks a lot! Signing out :)

<!-- gh-comment-id:2358257811 --> @KhazAkar commented on GitHub (Sep 18, 2024): I know that ollama won't be able to do so, it's understandable :) I've asked in internal communication if somebody have cURL way of negotiating captive portal :) Thanks a lot! Signing out :)
GiteaMirror commented 2026-04-12 15:08:31 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 18, 2024):

If it's just a button press, you could write a simple script run out of crontab to just hit the button every 30 minutes with whatever arguments are required to grant access.

<!-- gh-comment-id:2358259945 --> @rick-github commented on GitHub (Sep 18, 2024): If it's just a button press, you could write a simple script run out of crontab to just hit the button every 30 minutes with whatever arguments are required to grant access.
GiteaMirror commented 2026-04-12 15:08:31 -05:00
Author
Owner
Copy Link

@KhazAkar commented on GitHub (Sep 18, 2024):

If it's just a button press, you could write a simple script run out of crontab to just hit the button every 30 minutes with whatever arguments are required to grant access.

Will try to write such script, now with ollama help :D
For now I just installed lynx browser, enter any url, accept cookie and works fine for now.

<!-- gh-comment-id:2358268774 --> @KhazAkar commented on GitHub (Sep 18, 2024): > If it's just a button press, you could write a simple script run out of crontab to just hit the button every 30 minutes with whatever arguments are required to grant access. Will try to write such script, now with ollama help :D For now I just installed lynx browser, enter any url, accept cookie and works fine for now.
GiteaMirror commented 2026-04-12 15:08:32 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Sep 18, 2024):

@KhazAkar it seems like your issue has been resolved. Since OP hasn't responded in some time, I'm going to close this issue.

<!-- gh-comment-id:2359289196 --> @mxyng commented on GitHub (Sep 18, 2024): @KhazAkar it seems like your issue has been resolved. Since OP hasn't responded in some time, I'm going to close this issue.
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#4205
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?