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

[GH-ISSUE #1869] Provide instructions for running Ollama as a service in a GitLab CI/CD job #63108

New Issue
Closed
opened 2026-05-03 12:07:55 -05:00 by GiteaMirror · 11 comments
GiteaMirror commented 2026-05-03 12:07:55 -05:00
Owner
Copy Link

Originally created by @Bengt on GitHub (Jan 9, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/1869

I have an GitLab CI/CD job that runs my tests like this:

run-pytest-python3.11:
    needs:
        -   build-pytest-python3.11
    coverage: '/(?i)total.*? (100(?:\.0+)?\%|[1-9]?\d(?:\.\d+)?\%)$/'
    image:
        # yamllint disable-line rule:line-length
        name: registry.gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot:pytest-python3.11-latest
        entrypoint: [""]
    script:
        -   cd /1-llm-chatbot
        -   PYTHONPATH=. venv/bin/python -m pytest --cov=llm_chatbot

My test suite includes tests of the code that uses the Ollama server to answer the test requests. How do I run ollama serve as a GitLab service?

Originally created by @Bengt on GitHub (Jan 9, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/1869 I have an GitLab CI/CD job that runs my tests like this: ```yaml run-pytest-python3.11: needs: - build-pytest-python3.11 coverage: '/(?i)total.*? (100(?:\.0+)?\%|[1-9]?\d(?:\.\d+)?\%)$/' image: # yamllint disable-line rule:line-length name: registry.gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot:pytest-python3.11-latest entrypoint: [""] script: - cd /1-llm-chatbot - PYTHONPATH=. venv/bin/python -m pytest --cov=llm_chatbot ``` My test suite includes tests of the code that uses the Ollama server to answer the test requests. How do I run `ollama serve` as a [GitLab service](https://docs.gitlab.com/ee/ci/services/)?
GiteaMirror commented 2026-05-03 12:08:06 -05:00
Author
Owner
Copy Link

@aliirz commented on GitHub (Jan 9, 2024):

Have you considered running Ollama as a docker container and then exposing it as GitLab service?

<!-- gh-comment-id:1883374894 --> @aliirz commented on GitHub (Jan 9, 2024): Have you considered running Ollama as a docker container and then exposing it as GitLab service?
GiteaMirror commented 2026-05-03 12:08:08 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Jan 9, 2024):

+1 to running as a GitLab service.

Here's an (unvetted) example:

my-job:
  services:
    - name: ollama/ollama:0.1.19
      alias: ollama
  script:
    - nc -vz ollama 11434
<!-- gh-comment-id:1883483409 --> @mxyng commented on GitHub (Jan 9, 2024): +1 to running as a GitLab service. Here's an (unvetted) example: ```yaml my-job: services: - name: ollama/ollama:0.1.19 alias: ollama script: - nc -vz ollama 11434 ```
GiteaMirror commented 2026-05-03 12:08:10 -05:00
Author
Owner
Copy Link

@Bengt commented on GitHub (Jan 10, 2024):

When I try ollama pull, I get the following error:

Error: could not connect to ollama server, run 'ollama serve' to start it
<!-- gh-comment-id:1884686855 --> @Bengt commented on GitHub (Jan 10, 2024): When I try `ollama pull`, I get the following error: ``` Error: could not connect to ollama server, run 'ollama serve' to start it ```
GiteaMirror commented 2026-05-03 12:08:10 -05:00
Author
Owner
Copy Link

@Bengt commented on GitHub (Jan 10, 2024):

I have added an entrypoint to the ollama service like so, but that does not help, either:

services:
    -   alias: ollama
        name: ollama/ollama:0.1.19
        entrypoint: ["ollama", "serve"]
<!-- gh-comment-id:1884732517 --> @Bengt commented on GitHub (Jan 10, 2024): I have added an entrypoint to the ollama service like so, but that does not help, either: ``` services: - alias: ollama name: ollama/ollama:0.1.19 entrypoint: ["ollama", "serve"] ```
GiteaMirror commented 2026-05-03 12:08:11 -05:00
Author
Owner
Copy Link

@Bengt commented on GitHub (Jan 10, 2024):

Setting OLLAMA_HOSTdid not help, either.

<!-- gh-comment-id:1885465543 --> @Bengt commented on GitHub (Jan 10, 2024): Setting `OLLAMA_HOST`did not help, either.
GiteaMirror commented 2026-05-03 12:08:12 -05:00
Author
Owner
Copy Link

@Bengt commented on GitHub (Jan 10, 2024):

Activating the service logging like so:

https://docs.gitlab.com/ee/ci/services/#capturing-service-container-logs

variables:
  CI_DEBUG_SERVICES: "true"

Gives this:


Waiting for services to be up and running (timeout 30 seconds)...
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.835151997Z Couldn't find '/root/.ollama/id_ed25519'. Generating new private key.
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837068506Z Your new public key is: 
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837079866Z 
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837084625Z ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIOzRULmG6k0cYtK68EBuTNrFNw0CvVI/sVR8XqaHSj5F
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837088825Z 
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838342722Z 2024/01/10 19:38:24 images.go:808: total blobs: 0
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838445250Z 2024/01/10 19:38:24 images.go:815: total unused blobs removed: 0
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838710697Z 2024/01/10 19:38:24 routes.go:930: Listening on [::]:[11](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L11)434 (version 0.1.19)
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495087885Z 2024/01/10 19:38:25 shim_ext_server.go:[14](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L14)2: Dynamic LLM variants [cuda]
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495117114Z 2024/01/10 19:38:25 gpu.go:35: Detecting GPU type
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495682538Z 2024/01/10 19:38:25 gpu.go:40: CUDA not detected: Unable to load libnvidia-ml.so library to query for Nvidia GPUs: /usr/lib/wsl/lib/libnvidia-ml.so.1: cannot open shared object file: No such file or directory
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495820096Z 2024/01/10 19:38:25 gpu.go:46: ROCm not detected: Unable to load librocm_smi64.so library to query for Radeon GPUs: /opt/rocm/lib/librocm_smi64.so: cannot open shared object file: No such file or directory
[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495838[16](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L16)6Z 2024/01/10 [19](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L19):38:25 routes.go:953: no GPU detected
<!-- gh-comment-id:1885602046 --> @Bengt commented on GitHub (Jan 10, 2024): Activating the service logging like so: https://docs.gitlab.com/ee/ci/services/#capturing-service-container-logs ``` variables: CI_DEBUG_SERVICES: "true" ``` Gives this: ``` Waiting for services to be up and running (timeout 30 seconds)... [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.835151997Z Couldn't find '/root/.ollama/id_ed25519'. Generating new private key. [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837068506Z Your new public key is: [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837079866Z [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837084625Z ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIOzRULmG6k0cYtK68EBuTNrFNw0CvVI/sVR8XqaHSj5F [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.837088825Z [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838342722Z 2024/01/10 19:38:24 images.go:808: total blobs: 0 [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838445250Z 2024/01/10 19:38:24 images.go:815: total unused blobs removed: 0 [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838710697Z 2024/01/10 19:38:24 routes.go:930: Listening on [::]:[11](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L11)434 (version 0.1.19) [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495087885Z 2024/01/10 19:38:25 shim_ext_server.go:[14](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L14)2: Dynamic LLM variants [cuda] [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495117114Z 2024/01/10 19:38:25 gpu.go:35: Detecting GPU type [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495682538Z 2024/01/10 19:38:25 gpu.go:40: CUDA not detected: Unable to load libnvidia-ml.so library to query for Nvidia GPUs: /usr/lib/wsl/lib/libnvidia-ml.so.1: cannot open shared object file: No such file or directory [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495820096Z 2024/01/10 19:38:25 gpu.go:46: ROCm not detected: Unable to load librocm_smi64.so library to query for Radeon GPUs: /opt/rocm/lib/librocm_smi64.so: cannot open shared object file: No such file or directory [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:25.495838[16](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L16)6Z 2024/01/10 [19](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L19):38:25 routes.go:953: no GPU detected ```
GiteaMirror commented 2026-05-03 12:08:13 -05:00
Author
Owner
Copy Link

@Bengt commented on GitHub (Jan 10, 2024):

Do I maybe need to configure the web origin hosts?

https://github.com/jmorganca/ollama/blob/main/docs/faq.md#how-can-i-allow-additional-web-origins-to-access-ollama

<!-- gh-comment-id:1885754820 --> @Bengt commented on GitHub (Jan 10, 2024): Do I maybe need to configure the web origin hosts? https://github.com/jmorganca/ollama/blob/main/docs/faq.md#how-can-i-allow-additional-web-origins-to-access-ollama
GiteaMirror commented 2026-05-03 12:08:13 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Jan 10, 2024):

Logs indicate the service is up and running and serving 0.0.0.0:

[service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838710697Z 2024/01/10 19:38:24 routes.go:930: Listening on [::]:[11](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L11)434 (version 0.1.19)

Keep in mind the address GitLab exposes is alias:port so OLLAMA_HOST must be set for the client like this OLLAMA_HOST=ollama:11434 ollama pull. While the port (11434) should be exposed by default, it's possible GitLab requires it to be set explicitly.

<!-- gh-comment-id:1885774089 --> @mxyng commented on GitHub (Jan 10, 2024): Logs indicate the service is up and running and serving 0.0.0.0: ``` [service:ollama__ollama-ollama-ollama-ollama] 2024-01-10T19:38:24.838710697Z 2024/01/10 19:38:24 routes.go:930: Listening on [::]:[11](https://gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot/-/jobs/5904376120#L11)434 (version 0.1.19) ``` Keep in mind the address GitLab exposes is `alias:port` so OLLAMA_HOST must be set for the client like this `OLLAMA_HOST=ollama:11434 ollama pull`. While the port (11434) should be exposed by default, it's possible GitLab requires it to be set explicitly.
GiteaMirror commented 2026-05-03 12:08:14 -05:00
Author
Owner
Copy Link

@Bengt commented on GitHub (Jan 11, 2024):

I couldn't make the service work, so for now, I am managing the Ollama server in a bash process like so:

pytest-python3.11:
    before_script:
        -   apt update
        -   apt install --yes curl
        -   curl https://ollama.ai/install.sh | sh
    coverage: '/(?i)total.*? (100(?:\.0+)?\%|[1-9]?\d(?:\.\d+)?\%)$/'
    image:
        # yamllint disable-line rule:line-length
        name: registry.gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot:pytest-python3.11-latest
        entrypoint: [""]
    script:
        -   ollama serve &
        -   pid=$!
        -   export OLLAMA_HOST=localhost
        -   export PYTHONPATH=.
        -   cd /1-llm-chatbot
        #   Wait for ollama to start
        -   sleep 60
        -   ollama pull llama2
        -   venv/bin/python -m pytest --cov=llm_chatbot
        -   kill $pid
<!-- gh-comment-id:1886006561 --> @Bengt commented on GitHub (Jan 11, 2024): I couldn't make the service work, so for now, I am managing the Ollama server in a bash process like so: ``` pytest-python3.11: before_script: - apt update - apt install --yes curl - curl https://ollama.ai/install.sh | sh coverage: '/(?i)total.*? (100(?:\.0+)?\%|[1-9]?\d(?:\.\d+)?\%)$/' image: # yamllint disable-line rule:line-length name: registry.gitlab.com/openknowledge-gmbh/projects/ml-platform/1-llm-chatbot:pytest-python3.11-latest entrypoint: [""] script: - ollama serve & - pid=$! - export OLLAMA_HOST=localhost - export PYTHONPATH=. - cd /1-llm-chatbot # Wait for ollama to start - sleep 60 - ollama pull llama2 - venv/bin/python -m pytest --cov=llm_chatbot - kill $pid ```
GiteaMirror commented 2026-05-03 12:08:15 -05:00
Author
Owner
Copy Link

@Bengt commented on GitHub (Jan 11, 2024):

Thanks for the hint, @mxyng! I think, I did that at some point, but maybe I was mistaking. I will try that again, later.

<!-- gh-comment-id:1886007453 --> @Bengt commented on GitHub (Jan 11, 2024): Thanks for the hint, @mxyng! I think, I did that at some point, but maybe I was mistaking. I will try that again, later.
GiteaMirror commented 2026-05-03 12:08:16 -05:00
Author
Owner
Copy Link

@DevOpsJeremy commented on GitHub (Feb 1, 2025):

This is the solution I came up with. Run ollama serve as a service, then pull the model via the Ollama API in the before_script:

ollama:
  services:
    - name: ollama/ollama
      alias: ollama
      command: ["serve"]
  before_script:
    - curl ollama:11434/api/pull -sS -X POST -d "{\"model\":\"llama3.2\",\"stream\":false}"
  script: curl -sS ollama:11434
<!-- gh-comment-id:2629090588 --> @DevOpsJeremy commented on GitHub (Feb 1, 2025): This is the solution I came up with. Run `ollama serve` as a service, then pull the model via the Ollama API in the `before_script`: ```yaml ollama: services: - name: ollama/ollama alias: ollama command: ["serve"] before_script: - curl ollama:11434/api/pull -sS -X POST -d "{\"model\":\"llama3.2\",\"stream\":false}" script: curl -sS ollama:11434 ```
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#63108
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?