Examples: query, "exact match", wildcard*, wild?ard, wild*rd
Fuzzy search: cake~ (finds cakes, bake)
Term boost: "red velvet"^4, chocolate^2
Field grouping: tags:(+work -"fun-stuff")
Escaping: Escape characters +-&|!(){}[]^"~*?:\ with \, e.g. \+
Range search: properties.timestamp:[1587729413488 TO *] (inclusive), properties.title:{A TO Z}(excluding A and Z)
Combinations: chocolate AND vanilla, chocolate OR vanilla, (chocolate OR vanilla) NOT "vanilla pudding"
Field search: properties.title:"The Title" AND text
Answered
Hi All We Have Set Nginx In Front Of Clearml And Signed With Our Own Self-Signed Certs I'M Trying To Modify The

Hi all
we have set nginx in front of ClearML and signed with our own self-signed certs

I'm trying to modify the cleanup_service to fit some of our needs (running remotely, using the SDK)
when trying to use the APIClient to delete task using from clearml.backend_api.session.client import APIClient )

warning errors which I can leave with:

InsecureRequestWarning: Certificate verification is disabled! Adding certificate verification is strongly advised. See: 

but when it tries to delete artifacts it fails with:

Retrying (Retry(total=2, connect=5, read=5, redirect=5, status=None)) after connection broken by 'SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1000)'))': /my-root-project/lions/clearml_try/try_2023-01-29_05-46-24.1b6ef1f57ee8463aace328e806a0b55b/metrics/cfg_0/Model/cfg_0_Model_00000000.jpeg

is there a way to tell tell the APIClient

  • skip the ssl_verify (already have api.verify_certificate : false in my ~/clearml.conf)
  • point my conda env to the CA located on the workstation I interact with ClearML backend
  
  
Posted 8 months ago
Votes Newest

Answers 15


@<1523701435869433856:profile|SmugDolphin23> thanks for good pointers

it did not work on first attempt - requests did not validated the certs right
I have added this:

token_req = requests.get(api_server + "/auth.login", verify="<my_org_CA>", auth=(access_key, secret_key))``` 
print(token_req)

I got back

<Response [200]>

which I believe is good right?

when adding
token = token_req.json()["data"]["token"]

I got errors from json decoder, which I believe is expected?

requests.exceptions.JSONDecodeError: Expecting value: line 1 column 1 (char 0)
  
  
Posted 8 months ago

Yes it should word with ClearML if it works with requests

  
  
Posted 8 months ago

looks like I can't interact with fileserver ?

  
  
Posted 8 months ago

SDK version: 1.14.4
clearml-server version: Server: 1.14.0-431 • API: 2.28

  
  
Posted 8 months ago

I'm using rpm based machine, but I get your direction
put the cert in the right place for python to looks for it automatically

can I assume if it works smoothly with requests or urllib3 it will work for the ClearML API?

  
  
Posted 8 months ago

so I think I'm in the right direction
adding verify= and pointing to my CA.pem looks like the right approach

now, how do I use it with ClearML API?

cleanup_service

for task in tasks:
    try:
      deleted_task = Task.get_task(task_id=task.id)
      print (deleted_task.name)
      deleted_task.delete(
               delete_artifacts_and_models=True,
               skip_models_used_by_other_tasks=True,
               raise_on_error=False
      )

it throw down the SSL error, for each that have stuff in fileserver that it's trying to delete

when uploading anything to fileserver

    task = Task.init(project_name=project_name, task_name=exp_name, continue_last_task=True)
    logger = task.get_logger()

    img = Image.open(f"./1_model.png").convert("RGB")

    logger.report_image(title=f"cfg_0", series="Model", iteration=1, image=img)
  
  
Posted 8 months ago

Hi @<1523701842515595264:profile|PleasantOwl46> , can you check what this results in?

from clearml.storage.helper import StorageHelper
StorageHelper.get(Task._get_default_session().get_files_server_host())._container.session.verify
  
  
Posted 8 months ago

I have tried some small task only uploads single file

logger = task.get_logger()
img = Image.open(f"./1_model.png").convert("RGB")
logger.report_image(title=f"cfg_0", series="Model", iteration=1, image=img)

ended with:

Retrying (Retry(total=0, connect=5, read=5, redirect=5, status=None)) after connection broken by 'SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1006)'))': /
2024-03-03 09:53:54,987 - clearml.metrics - WARNING - Failed uploading to 
 (HTTPSConnectionPool(host='clearml-hrl-03.haifa.ibm.com', port=8081): Max retries exceeded with url: / (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1006)'))))
2024-03-03 09:53:54,988 - clearml.metrics - ERROR - Not uploading 1/1 events because the data upload failed
  
  
Posted 8 months ago

Hi @<1523701842515595264:profile|PleasantOwl46> ! This looks like a python problem. A useful SO thread: None
First, I would verify that I can access the api server without using the SDK. To do so, run this code after filling the credentials yourself (just login should be enough to verify that the api server is reachable)

    api_server = ""
    access_key = ""
    secret_key = ""
    token_req = requests.get(api_server + "/auth.login", auth=(access_key, secret_key))
    token = token_req.json()["data"]["token"]

If it doesn't work, it means that requests can't validate the certificates properly.
If that is the case, then I would try to install certifi both via pip install -U certifi and apt install -y python3-certifi (commands might vary) and the try to do make the get with the argument verifi=certifi.where() . Then setting REQUESTS_CA_BUNDLE env var to the right certifi path should allow the requests to go through.
If this also doesn't work, then I would look for the right certificates, because some comodo certificates for example may not be downloaded via certifi: (for example: None ).

These are just a few suggestions, not sure if they will help

  
  
Posted 8 months ago

have you tried copying the certificate to /usr/local/share/ca-certificates/ ?

  
  
Posted 8 months ago

@<1523701087100473344:profile|SuccessfulKoala55> looks OK (?)

>>> StorageHelper.get(Task._get_default_session().get_files_server_host())._container.session.verify
InsecureRequestWarning: Certificate verification is disabled! Adding certificate verification is strongly advised. See: 

True
  
  
Posted 8 months ago

the sudo update-ca-certificates ? maybe this will work

  
  
Posted 8 months ago

let me dig in more and hopefully can share successful results
thanks!

  
  
Posted 8 months ago

in case this will help someone else, I did not had root access to the training machine to add the cert to store
you can point your python to your own CA using:

export CURL_CA_BUNDLE=/path/to/CA.pem
  
  
Posted 8 months ago

@<1523701435869433856:profile|SmugDolphin23> working! here is what I have on Fedora/RHEL

  • copy certs to /etc/pki/ca-trust/source/anchors/
  • update-ca-trust
  
  
Posted 8 months ago
646 Views
15 Answers
8 months ago
8 months ago
Tags