feat: update jupyter-tensorflow-full rock for 1.8

[NohaIhab]

Update the jupyter-tensorflow-full rock for 1.8, based on the upstream image

Summary of changes

• change the branch to v1.8-branch
• bump the rock version to 1.8
• pin pyjuju to <4.0 in the unit test requirements due to migration to juju 3 in 1.8
• update kubectl to 1.25

Testing

Note: the rock was built before running the test

tox -e unit
unit installed: anyio==4.0.0,asttokens==2.4.0,attrs==23.1.0,backcall==0.2.0,bcrypt==4.0.1,cachetools==5.3.1,certifi==2023.7.22,cffi==1.15.1,charmed-kubeflow-chisme==0.2.0,charset-normalizer==3.2.0,cryptography==41.0.4,decorator==5.1.1,deepdiff==6.2.1,exceptiongroup==1.1.3,executing==1.2.0,google-auth==2.23.0,h11==0.14.0,httpcore==0.18.0,httpx==0.25.0,hvac==1.2.1,idna==3.4,importlib-resources==6.1.0,iniconfig==2.0.0,ipdb==0.13.13,ipython==8.12.2,jedi==0.19.0,Jinja2==3.1.2,jsonschema==4.17.3,juju==3.2.2,kubernetes==28.1.0,lightkube==0.14.0,lightkube-models==1.28.1.4,macaroonbakery==1.3.1,MarkupSafe==2.1.3,matplotlib-inline==0.1.6,mypy-extensions==1.0.0,oauthlib==3.2.2,ops==2.6.0,ordered-set==4.1.0,packaging==23.1,paramiko==2.12.0,parso==0.8.3,pexpect==4.8.0,pickleshare==0.7.5,pkgutil-resolve-name==1.3.10,pluggy==1.3.0,prompt-toolkit==3.0.39,protobuf==3.20.3,ptyprocess==0.7.0,pure-eval==0.2.2,pyasn1==0.5.0,pyasn1-modules==0.3.0,pycparser==2.21,Pygments==2.16.1,pyhcl==0.4.5,pymacaroons==0.13.0,PyNaCl==1.5.0,pyRFC3339==1.1,pyrsistent==0.19.3,pytest==7.4.2,pytest-asyncio==0.21.1,pytest-operator==0.29.0,python-dateutil==2.8.2,pytz==2023.3.post1,PyYAML==6.0.1,requests==2.31.0,requests-oauthlib==1.3.1,rsa==4.9,ruamel.yaml==0.17.32,ruamel.yaml.clib==0.2.7,serialized-data-interface==0.7.0,six==1.16.0,sniffio==1.3.0,stack-data==0.6.2,tenacity==8.2.3,tomli==2.0.1,toposort==1.10,traitlets==5.10.0,typing-extensions==4.8.0,typing-inspect==0.9.0,urllib3==1.26.16,wcwidth==0.2.6,websocket-client==1.6.3,websockets==8.1,zipp==3.17.0
unit run-test-pre: PYTHONHASHSEED='4273105998'
unit run-test: commands[0] | pytest -v --tb native --show-capture=all --log-cli-level=INFO /home/ubuntu/kubeflow-rocks/jupyter-tensorflow-full/tests
====================================================================================== test session starts ======================================================================================
platform linux -- Python 3.8.10, pytest-7.4.2, pluggy-1.3.0 -- /home/ubuntu/kubeflow-rocks/jupyter-tensorflow-full/.tox/unit/bin/python
cachedir: .tox/unit/.pytest_cache
rootdir: /home/ubuntu/kubeflow-rocks/jupyter-tensorflow-full
plugins: anyio-4.0.0, asyncio-0.21.1, operator-0.29.0
asyncio: mode=strict
collected 1 item                                                                                                                                                                                

tests/test_rock.py::test_rock 
---------------------------------------------------------------------------------------- live log setup -----------------------------------------------------------------------------------------
INFO     pytest_operator.plugin:plugin.py:647 Adding model microk8s-localhost:test-rock-e1rg on cloud microk8s
PASSED                                                                                                                                                                                    [100%]
--------------------------------------------------------------------------------------- live log teardown ---------------------------------------------------------------------------------------
INFO     pytest_operator.plugin:plugin.py:783 Model status:

Model           Controller          Cloud/Region        Version  SLA          Timestamp
test-rock-e1rg  microk8s-localhost  microk8s/localhost  3.1.5    unsupported  09:49:55Z

INFO     pytest_operator.plugin:plugin.py:789 Juju error logs:


INFO     pytest_operator.plugin:plugin.py:877 Resetting model test-rock-e1rg...
INFO     pytest_operator.plugin:plugin.py:882 Not waiting on reset to complete.
INFO     pytest_operator.plugin:plugin.py:855 Forgetting main...


====================================================================================== 1 passed in 10.21s =======================================================================================
unit run-test: commands[1] | python /home/ubuntu/kubeflow-rocks/jupyter-tensorflow-full/tests/test_access.py
Running jupyter-tensorflow-full_v1.8.0_20.04_1_amd64:v1.8.0_20.04_1
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  3914  100  3914    0     0   254k      0 --:--:-- --:--:-- --:--:--  254k
84ab5d20dbeb
84ab5d20dbeb
____________________________________________________________________________________________ summary ____________________________________________________________________________________________
  unit: commands succeeded
  congratulations :)

[i-chvets]

Are we going to add tenacity here instead of sleep?

[i-chvets]

@NohaIhab We need to update comments to indicate new branch:

# Based on the following Dockerfiles:
# https://github.com/kubeflow/kubeflow/blob/v1.7-branch/components/example-notebook-servers/base/Dockerfile
# https://github.com/kubeflow/kubeflow/blob/v1.7-branch/components/example-notebook-servers/jupyter/Dockerfile
# https://github.com/kubeflow/kubeflow/blob/v1.7-branch/components/example-notebook-servers/jupyter-tensorflow-full/cpu.Dockerfile

[i-chvets]

We are hitting incompatibility issues in our Jupyter Tensorflow ROCK. kfp* is pinned in upstream (has been for 6 months):

kfp==1.6.3
kfp-server-api==1.6.0

This is causing issues with urllib3:

ImportError: cannot import name 'appengine' from 'urllib3.contrib'

Looks like issue is fixed in kfp-1.8.21 and kfp-2.0.0b16

Options:

• Bump version to either 1.8.21 or 2.0.0b16 (dangerous). Will require us to have our own patch for ROCK (as per our spec) and create PR upstream.

[i-chvets]

Upstream v1.8-branch Docker is building with errors (container image is created successfully):

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
jupyterlab-server 2.25.0 requires jsonschema>=4.18.0, but you have jsonschema 3.2.0 which is incompatible.
tensorflow 2.9.3 requires absl-py>=1.0.0, but you have absl-py 0.11.0 which is incompatible.

[i-chvets]

Added imports tests with commented out (TO-DO) kfp import which causes dependency issue. This is the same way other ROCKs are handling this issue.

Created issue to track it: #53