mlflow: [BUG] UI not loading in the browser

Willingness to contribute

No. I cannot contribute a bug fix at this time.

MLflow version

2.7.1

System information

Windows 10 Python 3.8.5

Describe the problem

I’ve got a completely local configuration set up with an sqlite backend store, which has been working perfectly well for months. Since yesterday, I’m not able to view anything in the UI in the browser. The server itself is still working, because I can access it programmatically from my python scripts just as before.

The browser is showing the “Something went wrong” screen, see attached.

I need to point out that this problem started when I tried to configure tracking via a remote server, being Azure Machine Learning, and I was playing around with the tracking URI. Regrettably, I didn’t stop my local server while I was experimenting with this, so I suspect this has caused my local setup to not work properly, although as I mentioned above, programmatically I can still do everything, it’s only the UI that is a problem.

Other points to note:

My environment variable MLFLOW_TRACKING_URI is still showing the correct value (http://localhost:5000)
I’ve tried clearing the browser cache etc.
I’ve loaded the sqlite DB in a separate DB viewer and the data looks fine

Steps to reproduce the bug

I guess you would reproduce the bug by setting up a local config with sqlite and then modify the tracking URI to remote server while the local server is still running…

This is how I launch my local server:

mlflow server --backend-store-uri sqlite:///mlflow.db --default-artifact-root ./artifacts --host 0.0.0.0

And the address I put into the browser is:

http://localhost:5000

Code to generate data required to reproduce the bug


import mlflow
mlflow.set_tracking_uri('azureml://....')
mlflow.set_experiment('test_azureml_mlflow')

Is the console panel in DevTools showing errors relevant to the bug?

:5000/ajax-api/2.0/preview/mlflow/experiments/list:1     Failed to load resource: the server responded with a status of 404 (NOT FOUND)
FetchUtils.js:78 Fetch failed:  Response
Xd @ FetchUtils.js:78
RequestStateWrapper.js:93 ERROR Array(1)
aV @ RequestStateWrapper.js:93
react-dom.production.min.js:216 Error: A request error occurred.
    at aV (RequestStateWrapper.js:94:9)
    at n.value (RequestStateWrapper.js:77:9)
    at n.value (RequestStateWrapper.js:87:17)
    at Ua (react-dom.production.min.js:187:188)
    at za (react-dom.production.min.js:186:173)
    at Vs (react-dom.production.min.js:269:427)
    at kl (react-dom.production.min.js:250:347)
    at El (react-dom.production.min.js:250:278)
    at Sl (react-dom.production.min.js:250:138)
    at ml (react-dom.production.min.js:243:163)
as @ react-dom.production.min.js:216
AppErrorBoundary.js:19 Error: A request error occurred.
    at aV (RequestStateWrapper.js:94:9)
    at n.value (RequestStateWrapper.js:77:9)
    at n.value (RequestStateWrapper.js:87:17)
    at Ua (react-dom.production.min.js:187:188)
    at za (react-dom.production.min.js:186:173)
    at Vs (react-dom.production.min.js:269:427)
    at kl (react-dom.production.min.js:250:347)
    at El (react-dom.production.min.js:250:278)
    at Sl (react-dom.production.min.js:250:138)
    at ml (react-dom.production.min.js:243:163) Object
value @ AppErrorBoundary.js:19
contentScript.bundle.js:122 i18next: languageChanged en-GB
contentScript.bundle.js:122 i18next: initialized Object
:5000/#/:1 Uncaught (in promise) e
contentScript.bundle.js:103 Object

Does the network panel in DevTools contain failed requests relevant to the bug?

# Request URL
http://localhost:5000/ajax-api/2.0/preview/mlflow/experiments/list

# Status Code
404

# Remote Address
127.0.0.1:5000

# Response
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
<title>404 Not Found</title>
<h1>Not Found</h1>
<p>The requested URL was not found on the server. If you entered the URL manually please check your spelling and try again.</p>

# Request Headers
GET /ajax-api/2.0/preview/mlflow/experiments/list HTTP/1.1
Accept: */*
Accept-Encoding: gzip, deflate, br
Accept-Language: en-GB,en-US;q=0.9,en;q=0.8
Connection: keep-alive
Content-Type: application/json; charset=utf-8
Cookie: mp_49719c4dbff298b1e6c9adca527948b5_mixpanel=%7B%22distinct_id%22%3A%20%22%24device%3A18b023bc0af28a-027219d079f653-26031e51-1fa400-18b023bc0af28a%22%2C%22%24device_id%22%3A%20%2218b023bc0af28a-027219d079f653-26031e51-1fa400-18b023bc0af28a%22%2C%22%24initial_referrer%22%3A%20%22%24direct%22%2C%22%24initial_referring_domain%22%3A%20%22%24direct%22%7D
Host: localhost:5000
Referer: http://localhost:5000/
Sec-Fetch-Dest: empty
Sec-Fetch-Mode: cors
Sec-Fetch-Site: same-origin
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/117.0.0.0 Safari/537.36
sec-ch-ua: "Google Chrome";v="117", "Not;A=Brand";v="8", "Chromium";v="117"
sec-ch-ua-mobile: ?0
sec-ch-ua-platform: "Windows"