search

jeff1evesque / machine-learning

Web-interface + rest API for classification and regression (https://jeff1evesque.github.io/machine-learning.docs)
Other
258 stars 85 forks source link

#3271: Rancher AWS instance unhealthy #3272

Closed jeff1evesque closed 6 years ago

jeff1evesque commented 6 years ago

Resolves #3271.

jeff1evesque commented 6 years ago

Currently only the browserify container is unhealthy:

rancher

The following logs were obtained:

6/23/2018 6:37:37 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:37:39 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:37:40 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:37:52 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:38:07 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:38:22 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:38:24 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:38:37 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:38:52 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:39:07 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:39:22 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:39:37 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:39:52 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
6/23/2018 6:40:07 PMexec: "./entrypoint": stat ./entrypoint: no such file or directory
coveralls commented 6 years ago

Coverage Status

Coverage remained the same at 29.15% when pulling 45fe065d07b422c44a3a57c658ae583632f9aedb on bug-3271 into fe13cdb6daaa78138b2b80de6a9cff09126a070c on master.

jeff1evesque commented 6 years ago

Manually running our browserify container via ssh in the AWS environment runs without error:

root@ip-xxx-xxx-xxx-xxx:~/machine-learning/dockerfile# docker run --hostname browserify --name browserify -it jeff1evesque/ml-browserify:0.8
[...TRACE-OMITTED...]
  ./__tests__/layout/analysis.test.jsx
  ./__tests__/content/register.test.jsx
  ./__tests__/content/login.test.jsx
  ./content.jsx
done

> reactjs@1.0.0 build:browserify /var/machine-learning/src/jsx
> browserify -t [ babelify --presets env,stage-2,react ] content.jsx > ../../interface/static/js/content.js

> reactjs@1.0.0 watch:jsx /var/machine-learning/src/jsx
> onchange 'src/jsx/**/*.jsx' 'src/jsx/**/*.js' -- npm run build:browserify

^Croot@ip-172-31-47-17:~/machine-learning/dockerfile#

However, the install_rancher implementation, which calls upon the corresponding xxx-compose.[rancher].yml, fails with multiple No such file or directory on AWS:

browserify

jeff1evesque commented 6 years ago

One likely possibility, is a difference with the rancher implementation. Specifically, we'll need to take a closer look at our rancher-compose.yml, by checking the corresponding volumes mount.

jeff1evesque commented 6 years ago

Our browserify container now properly compiles:

browserify

Additionally, the corresponding compile asset is available on the AWS host:

root@ip-xxx-xxx-xxx-xxx:~/machine-learning# ls -l /root/machine-learning/interface/static/js
total 3116
-rw-r--r-- 1 root root 3190082 Jun 27 02:20 content.js
jeff1evesque commented 6 years ago

Our application is now accessible on AWS:

frontpage

However, some of our stylesheets have been incorrectly adjusted:

login

data-new