We asked Heroku support again and they're saying we should be able to use SELECT sum(size) FROM pg_ls_waldir(); To address your follow up question; internally we use a symlink from the pg_wal directory to a directory on a separate WAL drive. From there it gets pushed to an S3 bucket. We also wanted to get back to you about the tests you were running using the select sum(size) from pg_ls_waldir() function, and let you know that there can be discrepancies reported when comparing the results of the pg_ls_waldir() function against our server metrics (sample#wal-percentage-used) due to the nature of how they are both being measured internally, but ultimately using the pg_ls_waldir() function should be a good way to monitor WAL usage (considering the WAL disk is 68GB). We're going to test our script again and see if this measurement is comparable to the Heroku-provided one.

Presumably, every Post must be linked to a corresponding (no pun intended) User. This is what the Foreign Key is enforcing. It will not allow you to create a Post that contains a userId that does not match an existing User. You don't show a "seeder" for the Users. Is it as simple as you're trying to "seed" the Posts before "seeding" the Users?

A couple of comments on the code from what I see in the repel:Hold on. .idea♪ target and node_modules folders in the repository are a bad solution, because the size of the repository can be very fast above the available limits. A little justification can be found http://alblue.bandlem.com/2011/03/git-tip-of-week-ignoring-build-output.html ♪The use of untreated originators for any language is, in principle, considered to be a bad tone, as it can have a strong impact on the productivity of the application. The rationale and review of solutions can be found http://www.hanselman.com/blog/TheImportanceAndEaseOfMinifyingYourCSSAndJavaScriptAndOptimizingPNGsForYourBlogOrWebsite.aspx to optimize productivity. In your development, rely on the same Heroku to take pre-deploy steps, for example, maven/gulp of the project.Just because you've been relying on an unprocessed library reference code,very joke. If you check the contents node_modules in the branch masteryou won't find any files there that can't find your app.

Procfile should be without expansion .txt For example, you can keep the file in the notebook, the name of the file in the quote, the type-- all the files. ♪"

Heroku didn't realize you had a Python app. (sighs)Failed to detect set buildpackBecause he doesn't know how to pick him up and launch him, so he refuses to accept him like that.According to the https://devcenter.heroku.com/articles/python-support#activation If Heroku correctly recognises him, there must be a file at the root. requirements.txtat least empty (if there is no dependency):The Heroku Python Support will be applied to applications only when the application has a requirements.txt in the root directory. Even if an application has no module dependencies, it should include an empty requirements.txt to document that your app has no dependencies.The same is stated https://elements.heroku.com/buildpacks/heroku/heroku-buildpack-python :A requirements.txt file must be present at the root of your application's repository.

Heroku does not provide a storage space for downloadable files. You've only got 300ms for slug (prepared for launch) and a folder/tmp for kesha storage and fresh loaded files. This file is purified periodically, and your files loaded into it are missing.Fissile storage can be used https://aws.amazon.com/ru/s3/ by connecting him to a gem application fog or aws-s3♪ Look for S3 instructions through these hems with your file loader.carrierwave♪ paperclip)

When you pump your web application, the server carries out a number of standard tasks depending on the technology upstream. It's a collection of styles and scruples, minimization and compression of files, etc.Ultimately, the server receives a collected and launched application. With all styles, violets, images and other resources. It's a fucking app. slug♪A slug size - is the permissible size of such an annex. I mean, if your start-up app weighs more than 300 mb, the fuck won't let it go. So pay attention to the resources you're holding in the repository of the project. Archives, videos, heavy schedules - it's better if you don't have a code repository at all, but a separate vault. For example, http://aws.amazon.com/s3 ♪Access to the file system can give a shit, but for the storage of the temporary files, the kesha, the files on the apelad. There's no need to keep long-term files in that vault. Again, it's better to tie the external vault to the project.

I've been asking the question myself recently, and it's helped connect the library through the maven.<dependency> <groupId>com.github.pengrad</groupId> <artifactId>java-telegram-bot-api</artifactId> <version>2.1.1</version> </dependency> <repository> <id>jitpack.io</id> <url>https://jitpack.io</url> </repository>

Dyno type web or vapour worker+web Heroku https://devcenter.heroku.com/articles/free-dyno-hours free tariff....and raises new dyno when it comes to a request that takes some time, but lets Heroku not spend any of his resources on abandoned germ projects at all.Since Telegram &apos; s boots in the use of conventional API follow messages themselves, the telegram message will not lead to its awakening. Needed inlet Request, whereas API applications outgoing♪Probably if you change dyno s web ♪ worker (sighs) ProcfileNot adding webhe won't fall asleep. But it's not gonna be possible to service HTTP's queries, so the Express will be out. Botus doesn't care.Better get on the web. Then Telegram will be delivered through incoming HTTP queries that I'll be there. to wake the beard.As Heroku spends only 550 free machine-hours per month on a scout (a) 31 * 24 = 744(sing) a permanently working bot. ♪ Use everything, and the appreciable time will be seduced (it won't wake up). This limit could be extended to 1,000 (by 450 hours) with a bank card. Then he won't have to sleep if there are no other (natural) users in the same account.Information is received that a 24-hour application obligation at least 6 hours(24 - 6) * 31 = 558) That's it. inaccurate♪ Heroku https://blog.heroku.com/announcing_heroku_free_ssl_beta_and_flexible_dyno_hours#flexible-free-dyno-hours (Thanks for the tip https://ru.stackoverflow.com/users/182273/alex78191 (a) Clearly points to the possibility of keeping chalbot 24/7 free of charge.

Heroku is so arranged that your server will only be activated when it comes to treatment. If there's no contact to the site, your instans will turn off and all the information stored on it is compressed. If Heroku is used, it's better to send images to some long-term exposure storage facility (Swift, S3). All the more, they're almost all allowed to tie a domain name, so you can give img.yoursite.com images.

In my case, it was necessary to secure the keys properly:openssl req -newkey rsa:2048 -sha256 -nodes -keyout apache.key -x509 -days 365 -out apache.pem -subj "/C=RU/ST=Krasnodar/L=Krasnodar/O=./CN=IP address of my server available from outside."And correctly establish webhook:curl-F "url= https://ip:port/ "F "certificate=@apache.pem." https://api.telegram.org/botTOKEN/setWebhook

Problem solved by adding the line:passive =state true in the model file::ftp_servers => [ { :host => "site.com", :user => "user", :password => "password", :passive => true }

Yeah, Heroku works that way.He's got a ephemeric disk and can be recreated from the image at any time. That's when the app falls asleep, his disk is destroyed, and when the booths, it creates a new, separate machine. Therefore, nothing on the disk that is intended for long-term storage cannot be retained.We need to use the external vault. It is possible to keep it directly in the OBD system (they are not intended for this purpose, but in theory this is real), but it is possible to connect the external storage service with HTTP access like Amazon Simple Storage Service (S3) and to come up with something else.

Start in the terminal.heroku logs --tail Realtime.

It's nice, Nothing.♪ SQLite is not a web-based server, but a library to work with some locally available file.Heroku, class classic 12-factor cloud, operating ephemerical file systems♪ The cloud reserves the right to redesign the file system from time to time and your file with the SQLite database will therefore be periodically lost. With all the data.This was then done to make it easier for the annex to scale up, so that another copy of your application could be easily picked up, connected to the same databases via the network, and... almost double the number of overlapping requests. Only when SQLite begins to be treated by more than one client, it begins with a light presentation.So you have a hard choice either to give up. SQLiteor refuse Heroku♪...or somehow make the application store the database file in the external storage facility, but lifting the normal RSB with network access might be even easier.

The problem was in the evidence. length: { maximum: 2000 } I did. length: { minimum: 2000 }You always have to read the code carefully!

Throwed fifth line user.rballegedly a challenge to the method #provider Ooh. nil♪Apparently, auth relevant nil♪ How did that happen?In the first line SessionsController#create on site auth request.env['omniauth_auth']♪ ♪ ♪Yeah, there. nilthat key. env (average) no request exists. Key wrong.♪For omniauth_auth I need you. omniauth.auth♪♪ https://github.com/intridea/omniauth#integrating-omniauth-into-your-application def auth_hash request.env['omniauth.auth'] end

https://devcenter.heroku.com/articles/getting-started-with-ruby#introduction heroku login We need to bail on Heroku, and maybe you're missing the site stage for your decoration project on Heroku himself through the team. heroku create

I think we should try the ways in the Automatic collectstatic description. https://devcenter.heroku.com/articles/django-assets#automatic-collectstatic ♪ In fact, in the conclusion, build a log and says what to do:$ heroku run python manage.py collectstatic --noinput As a result, you'll get the information you need for the fix. For example, django.core.exceptions.ImproperlyConfigured: You're using the staticfiles app without having set the STATIC_ROOT setting.And you'll be able to better diagnose the source of the problem.UPD: As I wrote earlier, you've been listed in the slogan. django.core.exceptions.ImproperlyConfigured: You're using the staticfiles app without having set the STATIC_ROOT setting to a filesystem path.In settings.py, you should indicate the value for STATIC_ROOT, for example. STATIC_URL = '/static/' STATIC_ROOT = os.path.join(BASE_DIR, 'static') It may also happen that the /static/static folder will have to be created manually, but you will understand when STATIC_ROOT put it in settings.py and reset the app.

About what better I can't recommend. That's what Heroku writes about choosing Puma as an application server: https://devcenter.heroku.com/articles/ruby-default-web-server