Heroku Flask Selenium サーバー上だと上手くいかない

Herokuにサーバーを立ててFlaskのWebアプリを実行したのですが、
こちらの「Application error」というエラーが出てしまいます。

ローカルの環境だと何も問題なく動くのですが、
Herokuにでデプロイすると、なぜか上手くいきません。

何かHeroku上でSeleniumを使った時にある不具合とかあるのでしょうか？

今回実行した手順はこちらの方法を参考にしました。
https://qiita.com/nsuhara/items/76ae132734b7e2b352dd

errorLog
12021-12-13T17:10:16.986388+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
22021-12-13T17:10:16.986417+00:00 app[web.1]: OpenBLAS blas_thread_init: pthread_create failed for thread 6 of 8: Resource temporarily unavailable
32021-12-13T17:10:16.986429+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
42021-12-13T17:10:16.986458+00:00 app[web.1]: OpenBLAS blas_thread_init: pthread_create failed for thread 7 of 8: Resource temporarily unavailable
52021-12-13T17:10:16.986470+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
62021-12-13T17:10:17.088740+00:00 app[web.1]: [2021-12-13 17:10:17 +0000] [295] [INFO] Worker exiting (pid: 295)
72021-12-13T17:10:17.138312+00:00 app[web.1]: [2021-12-13 17:10:17 +0000] [4] [WARNING] Worker with pid 295 was terminated due to signal 11
82021-12-13T17:10:17.141040+00:00 app[web.1]: [2021-12-13 17:10:17 +0000] [298] [INFO] Booting worker with pid: 298
92021-12-13T17:10:17.281440+00:00 app[web.1]: OpenBLAS blas_thread_init: pthread_create failed for thread 3 of 8: Resource temporarily unavailable
102021-12-13T17:10:17.281467+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
112021-12-13T17:10:17.281517+00:00 app[web.1]: OpenBLAS blas_thread_init: pthread_create failed for thread 4 of 8: Resource temporarily unavailable
122021-12-13T17:10:17.281530+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
132021-12-13T17:10:17.281564+00:00 app[web.1]: OpenBLAS blas_thread_init: pthread_create failed for thread 5 of 8: Resource temporarily unavailable
142021-12-13T17:10:17.281577+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
152021-12-13T17:10:17.281609+00:00 app[web.1]: OpenBLAS blas_thread_init: pthread_create failed for thread 6 of 8: Resource temporarily unavailable
162021-12-13T17:10:17.281623+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
172021-12-13T17:10:17.281656+00:00 app[web.1]: OpenBLAS blas_thread_init: pthread_create failed for thread 7 of 8: Resource temporarily unavailable
182021-12-13T17:10:17.281669+00:00 app[web.1]: OpenBLAS blas_thread_init: RLIMIT_NPROC 256 current, 256 max
192021-12-13T17:10:17.384090+00:00 app[web.1]: [2021-12-13 17:10:17 +0000] [298] [INFO] Worker exiting (pid: 298)
202021-12-13T17:10:17.520105+00:00 app[web.1]: [2021-12-13 17:10:17 +0000] [4] [WARNING] Worker with pid 298 was terminated due to signal 11
212021-12-13T17:10:17.523608+00:00 app[web.1]: [2021-12-13 17:10:17 +0000] [301] [INFO] Booting worker with pid: 301
222021-12-13T17:42:46.730303+00:00 heroku[web.1]: Idling
232021-12-13T17:42:46.769331+00:00 heroku[web.1]: State changed from up to down
242021-12-13T17:42:47.662870+00:00 heroku[web.1]: Stopping all processes with SIGTERM
252021-12-13T17:42:47.737256+00:00 app[web.1]: [2021-12-13 17:42:47 +0000] [9] [INFO] Worker exiting (pid: 9)
262021-12-13T17:42:47.737268+00:00 app[web.1]: [2021-12-13 17:42:47 +0000] [301] [INFO] Worker exiting (pid: 301)
272021-12-13T17:42:47.739529+00:00 app[web.1]: [2021-12-13 17:42:47 +0000] [4] [INFO] Handling signal: term
282021-12-13T17:42:47.752851+00:00 app[web.1]: [2021-12-13 17:42:47 +0000] [4] [WARNING] Worker with pid 301 was terminated due to signal 15
292021-12-13T17:42:47.755111+00:00 app[web.1]: [2021-12-13 17:42:47 +0000] [4] [WARNING] Worker with pid 9 was terminated due to signal 15
302021-12-13T17:42:47.840939+00:00 app[web.1]: [2021-12-13 17:42:47 +0000] [4] [INFO] Shutting down: Master
312021-12-13T17:42:48.046917+00:00 heroku[web.1]: Process exited with status 0
322021-12-13T17:54:34.156378+00:00 heroku[web.1]: Unidling
332021-12-13T17:54:34.203324+00:00 heroku[web.1]: State changed from down to starting
342021-12-13T17:54:47.463452+00:00 heroku[web.1]: Starting process with command `gunicorn --bind 0.0.0.0:46382 root:app`
352021-12-13T17:54:48.941099+00:00 app[web.1]: [2021-12-13 17:54:48 +0000] [4] [INFO] Starting gunicorn 20.1.0
362021-12-13T17:54:48.941516+00:00 app[web.1]: [2021-12-13 17:54:48 +0000] [4] [INFO] Listening at: http://0.0.0.0:46382 (4)
372021-12-13T17:54:48.941562+00:00 app[web.1]: [2021-12-13 17:54:48 +0000] [4] [INFO] Using worker: sync
382021-12-13T17:54:48.947339+00:00 app[web.1]: [2021-12-13 17:54:48 +0000] [9] [INFO] Booting worker with pid: 9
392021-12-13T17:54:48.981241+00:00 app[web.1]: [2021-12-13 17:54:48 +0000] [10] [INFO] Booting worker with pid: 10
402021-12-13T17:54:49.326215+00:00 heroku[web.1]: State changed from starting to up
41

heroku logs --tailにてログを出してみましたが、
いまいちなぜHeroku上だとエラーが出るのか理解できていません。

詳しい方いましたらご教授願います。

行動規範の内容に同意します

回答1件

Possible fixes include:

explicitly setting the number of openBLAS threads using:
export OMP_NUM_THREADS=1
restricting the numpy version to be <1.16.0

https://github.com/bcgsc/mavis/issues/185#issuecomment-481376728
上ののissueについたコメントで言及されているように、 numpyのバージョンを下げるかOMP_NUM_THREADS=1で環境変数を設定したらエラーが消えて動くように思います。

投稿2021/12/14 02:37