kubeflow / katib

Automated Machine Learning on Kubernetes
https://www.kubeflow.org/docs/components/katib
Apache License 2.0
1.51k stars 443 forks source link

Katib Experiment Graph is not shown #1196

Closed sakaia closed 3 years ago

sakaia commented 4 years ago

/kind bug

What steps did you take and what happened: I can successfully execute the random-experiment but graph is not shown (empty). But I can see result of execution in text

katib katib1

What did you expect to happen: I can run the code with graph output. (item 6 Figure) example-using-random-algorithm

Anything else you would like to add:

Environment:

issue-label-bot[bot] commented 4 years ago

Issue-Label Bot is automatically applying the labels:

Label Probability
area/katib 0.98

Please mark this comment with :thumbsup: or :thumbsdown: to give our bot feedback! Links: app homepage, dashboard and code for this bot.

andreyvelich commented 4 years ago

@sakaia Can you check logs from katib-ui pod, please?

issue-label-bot[bot] commented 4 years ago

Issue-Label Bot is automatically applying the labels:

Label Probability
area/front-end 0.77

Please mark this comment with :thumbsup: or :thumbsdown: to give our bot feedback! Links: app homepage, dashboard and code for this bot.

sakaia commented 4 years ago

Thank you My katib-ui pod logs are follows. but, still graph is not shown.

kubectl logs katib-ui-6f97756598-gclzx -n kubeflow
2020/05/26 23:25:23 Serving the frontend dir /app/build
2020/05/26 23:25:23 Serving at 0.0.0.0:8080
2020/05/26 23:32:07 Got Experiment
2020/05/26 23:32:07 Got metrics names
2020/05/26 23:32:07 Got Parameters names
2020/05/26 23:32:07 Got Trial List
2020/05/26 23:32:07 Logs parsed, results:
 trialName,Status,Validation-accuracy,accuracy,--lr,--num-layers,--optimizer
2020/05/26 23:32:25 Got Experiment
2020/05/26 23:32:25 Got metrics names
2020/05/26 23:32:25 Got Parameters names
2020/05/26 23:32:25 Got Trial List
2020/05/26 23:32:25 Logs parsed, results:
 trialName,Status,Validation-accuracy,accuracy,--lr,--num-layers,--optimizer
2020/05/26 23:32:33 Got Experiment
2020/05/26 23:32:33 Got metrics names
2020/05/26 23:32:33 Got Parameters names
2020/05/26 23:32:33 Got Trial List
2020/05/26 23:32:33 Logs parsed, results:
 trialName,Status,Validation-accuracy,accuracy,--lr,--num-layers,--optimizer
random-experiment-9h7kxz48,Running,,,0.014115112699862925,5,ftrl
random-experiment-mq9rkr8q,Running,,,0.015907948511341558,2,ftrl
random-experiment-z9ck29hh,Running,,,0.013710131519663613,3,ftrl
2020/05/26 23:39:16 Got Experiment
2020/05/26 23:39:16 Got metrics names
2020/05/26 23:39:16 Got Parameters names
2020/05/26 23:39:16 Got Trial List
2020/05/26 23:39:16 Logs parsed, results:
 trialName,Status,Validation-accuracy,accuracy,--lr,--num-layers,--optimizer
random-experiment-8k4f75w9,Succeeded,0.113854,0.121719,0.024844551620860085,3,ftrl
random-experiment-9h7kxz48,Succeeded,0.113854,0.121906,0.014115112699862925,5,ftrl
random-experiment-9tfq8gcb,Succeeded,0.113854,0.120938,0.027901018829838475,2,ftrl
random-experiment-c4vl6rck,Succeeded,0.113854,0.122969,0.02976483919797447,3,ftrl
random-experiment-dxsqpqq6,Succeeded,0.978603,0.996250,0.017494586655338925,3,sgd
random-experiment-h295h7lc,Running,,,0.027933050705964572,2,sgd
random-experiment-ltdbrpwf,Running,,,0.025982093612760965,4,ftrl
random-experiment-mq9rkr8q,Succeeded,0.113854,0.124219,0.015907948511341558,2,ftrl
random-experiment-rskwmqc6,Succeeded,0.113854,0.122656,0.011650364554826508,2,ftrl
random-experiment-sknhmt9l,Succeeded,0.113854,0.123906,0.018498079446027316,4,ftrl
random-experiment-sp76xt94,Running,,,0.021721815138828454,4,adam
random-experiment-z9ck29hh,Succeeded,0.113854,0.125156,0.013710131519663613,3,ftrl
andreyvelich commented 4 years ago

@sakaia Thanks, logs are looking good.

Try to update this page few times and check errors inside the Google Chrome console. Also, try to click on any Trial name and see if Metrics Graph is shown.

sakaia commented 4 years ago

Thank you for your suggestion.

It seems following file problem (on jsx). Is there any good way to monitor it?

pkg/ui/v1alpha3/frontend/src/components/HP/Monitor/HPJobInfo.jsx

(react-ploty.js seems drwaring soft for graph?)

andreyvelich commented 4 years ago

@sakaia Yes, we are using react-plotly.js to draw this plot. I think you can check frontend errors in the browser console. Maybe you need to enable JS in your browser?

Or try to disable cache in the console and refresh page:

Screenshot 2020-05-27 at 17 55 10
sakaia commented 4 years ago

Thank you for your suggestion. I try to disable cache but the problem still remain. app bundle returns 503. Is there any way to digging this?

katibhpgithub
gaocegege commented 4 years ago

Can you show us the request detail of the 503 request?

sakaia commented 4 years ago

Is this view help to investigate?

katibgithub2
sakaia commented 4 years ago

PS. I can see nas graph. but I still cannot see HP graph.

katibnas1
andreyvelich commented 4 years ago

@sakaia Can you try to update your katib-ui deployment image to the latest. You don't need to redeploy all Kubeflow components, just run: kubectl edit deploy katib-ui -n kubeflow and modify the container image to: gcr.io/kubeflow-images-public/katib/v1alpha3/katib-ui. After that, the katib-ui pod will be restarting.

Then, try to access Katib UI using port-forwarding: kubectl port-forward svc/katib-ui 8080:80 -n kubeflow.

You should be able to access Katib UI with this URL: http://localhost:8080/katib. Try if it works for you.

sakaia commented 4 years ago

Thank you after updating from v0.8.0 to nothing. Web UI error is changed to follows. Even I enable WebGL on chrome, the problem still keeps

katibhpgithub2

My environment is follows.

Google Chrome 81.0.4044.138 (Official Build) (64 ビット)
変更履歴 8c6c7ba89cc9453625af54f11fd83179e23450fa-refs/branch-heads/4044@{#999}
OS Windows 8.1 (Build 9600)
JavaScript V8 8.1.307.32
Flash 24.0.0.189 internal-not-yet-present
ユーザー エージェント Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/81.0.4044.138 Safari/537.36
コマンドライン "C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --enable-audio-service-sandbox --flag-switches-begin --enable-webgl-draft-extensions --enable-webgl2-compute-context --flag-switches-end --enable-audio-service-sandbox
実行ファイルのパス C:\Program Files (x86)\Google\Chrome\Application\chrome.exe
andreyvelich commented 4 years ago

@sakaia Thanks, at least we can see the error now. Maybe try to disable --enable-webgl-draft-extensions --enable-webgl2-compute-context flags. Try to visit https://get.webgl.org/ to see if WebGL works in your browser after it.

stale[bot] commented 4 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

stale[bot] commented 3 years ago

This issue has been automatically closed because it has not had recent activity. Please comment "/reopen" to reopen it.