JSON request returns two responses - ruby-on-rails

In my rails app I have an Item model which has many Versions. My controller has a versions action.
def versions
#item = Item.find(params[:id])
respond_to do |format|
format.json { render json: #item.versions }
end
end
Whenever I request versions, some items return them successfully and some don't. But when I go into the console, I can get the versions of all items with no problem.
Successful (returns JSON array of versions):
Started GET "/api/versions/4.json?"
Processing by ItemsController#versions as JSON
Parameters: {"id"=>"4"}
Completed 200 OK in 106ms (Views: 0.1ms | ActiveRecord: 19.8ms)
app[web.1]: cache: [GET /api/versions/4.json?] miss
heroku[router]: GET x.x.com/api/versions/4.json? dyno=web.1 queue=0 wait=0ms service=113ms status=200 bytes=831
Not successful (returns two responses, both status 200, no JSON data):
Started GET "/api/versions/1.json?"
Processing by ItemsController#versions as JSON
Parameters: {"id"=>"1"}
Completed 200 OK in 249ms (Views: 0.1ms | ActiveRecord: 162.9ms)
app[web.1]: cache: [GET /api/versions/1.json?] miss
heroku[router]: GET x.x.com/api/versions/1.json? dyno=web.1 queue=0 wait=0ms service=257ms status=200 bytes=3540
Any idea what could cause this? I don't know how to log SQL on Heroku so I'm out of ideas.

Re your SQL logging: Enable statement logging by issuing the SQL SET log_statement='all'; when you open each connection. Then examine the Heroku logs for Pg to see what's happening.
heroku logs --ps postgres
See:
Heroku's documentation on their PostgreSQL support
Heroku's logging documentation
a Google search for: how to log SQL on Heroku
This won't log query results, only query text, query parameters, and error/success.
To handle result logging you'll need to do it in your application; just print messages to stderr and Heroku will add them to the application logs so you can get them via heroku logs.
Your current code looks dodgy anyway. Looping over a result set and emitting json feels wrong:
It isn't explicitly clear that there can only be one result and thus only emit one json block; and
If there are no results, no json is emitted
Instead, save the result of the find - if any - to a variable. Then handle missing results properly: in a RESTful API you'd probably report HTTP status 404 Not Found; in a more RPC-style API you'd return an empty JSON document or a JSON document with a "not found" flag of some kind and report HTTP status 200 OK.

This was actually an error on the client side. On iOS I was handling the returned data in connection:didReceiveData:, which (when returning large chunks of data) is called numerous times with partial results. To fix it, I instead concatenated all the partial results in that callback and handled the data in connectionDidFinishLoading:. Stupid mistake, not related to Heroku or Rails.
Thanks for the help Craig.

Related

axios/rack-cors/react-rails/heroku: Internal Server Error 500 on sign out, no persistence on refresh

ERROR MSG:
Error: Request failed with status code 500
Stack trace:
[42]/</t.exports#http://crdwk.herokuapp.com/packs/bundle-ecc8ea14dbe153e50352.js:1:89311
[42]/</t.exports#http://crdwk.herokuapp.com/packs/bundle-ecc8ea14dbe153e50352.js:1:251725
[42]/</t.exports/</d[h]#http://crdwk.herokuapp.com/packs/bundle-ecc8ea14dbe153e50352.js:1:88311
Ruby version: 2.3
Rails version: 5.1
I have a server-side rendered, client-side hydrated React/Rails app (using gem 'react-rails').
I added the gem 'rack-cors' plus setup in application.rb in order for my requests to work (I'm using axios). However, signing out (a DELETE request) fails and hitting refresh erases the current user. Neither issue occurs locally/in development.
Here's the app: http://crdwk.herokuapp.com
And the repo: https://github.com/English3000/crdwk
I took a look at my Heroku logs:
Completed 500 Internal Server Error in 8ms (ActiveRecord: 0.0ms)
NoMethodError (undefined method `reset_token' for nil:NilClass):
app/controllers/application_controller.rb:29:in `sign_out'
app/controllers/api/sessions_controller.rb:3:in `destroy'
Given that hitting refresh, the current user does not persist, the issue is the current user somehow isn't getting set.
However, this is not an issue in development. Why would that be?
Looking through my project, the only difference I can find on the backend as compared with a client-side rendered one (which I literally copy & pasted the code for this project from) is this line in application_controller.rb:
skip_before_action :verify_authenticity_token
However, if I comment out this line, when I try to sign up, I get the server error
Started POST "/api/users" for 127.0.0.1 at 2018-03-05 12:16:57 -0800
Processing by Api::UsersController#create as JSON
Parameters: {"user"=>{"email"=>"", "password"=>"[FILTERED]"}}
Can't verify CSRF token authenticity.
Completed 422 Unprocessable Entity in 1ms (ActiveRecord: 0.0ms)
ActionController::InvalidAuthenticityToken - ActionController::InvalidAuthenticityToken:
This is a result of using gem 'react-rails'. (I don't get this error for my other client-side rendered project.)
There isn't an "authenticity_token" parameter.
I find these two resources: Rails security and Learnetto's how-to.
So I add these two lines of code from the second artilce to my api.js:
const csrfToken = document.querySelector("meta[name=csrf-token]").content;
axios.defaults.headers.common["X-CSRF-Token"] = csrfToken;
Now my web app works with the extra application_controller.rb line commented out. EXCEPT, I can't use the DOM to grab the csrf token for my React Native version, so I now have the same issue for mobile...

Rails production.log (using Passenger) being written to by multiple processes, can't be parsed

Here's a snippet of my production.log:
Started GET "/product/514034/754240" for XX.XX.202.138 at 2012-06-21 11:52:28 -0700
Started GET "/product/614409/666897" for XX.XX.228.38 at 2012-06-21 11:52:28 -0700
Processing by ProductsController#show as HTML
Parameters: {"category_id"=>"514034", "product_id"=>"754240"}
Processing by ProductsController#show as HTML
Parameters: {"category_id"=>"614409", "product_id"=>"666897"}
Logged in 2940659 via auth cookie
Logged in 585210 via auth cookie
[e3e3fc56bb6bd137741b269ee397683c] [2940659] Read fragment views/global-caches/header (0.7ms)
[e3e3fc56bb6bd137741b269ee397683c] [2940659] Rendered shared/_email_form.html.haml (0.7ms)
[d81bb986be5acc0277c0c9e11b414249] [585210] Read fragment views/global-caches/sharebar-message (0.7ms)
[d81bb986be5acc0277c0c9e11b414249] [585210] Rendered shared/_email_form.html.haml (0.7ms)
...
As you can see, it's logging two concurrent sessions of two different users simultaneously to the same log file. This makes it impossible to parse my logs and determine, for example, the time it took to generate each kind of page, because the entries are not in the expected order of:
Started GET "/URL/BLAH" for IP at DATE
... stuff...
Completed 200 OK in XXms (ActiveRecord: YY.Yms)
Instead I get an unpredictable interleaved log like this:
Started GET "/URL/BLAH" for IP at DATE
Started GET "/URL/BLAH" for IP at DATE
... stuff...
Completed 200 OK in XXms (ActiveRecord: YY.Yms)
...stuff...
Completed 200 OK in XXms (ActiveRecord: YY.Yms)
So it's impossible to match the "completeds" with the "Started."
What I'd like is a way to have each child process write to its own log or something. Or if it's possible a way to write the each pageview's log atomically, but that might be impossible or difficult or hurt performance.
Rails 3.2 provides nice option config.log_tag
You can add to your production.rb:
config.log_tags = [ lambda { Time.now.to_i }]
So each line in your logs will be prepended by numbers. Example:
[1351867173] Started GET "/" for 127.0.0.1 at 2012-11-02 16:39:33 +0200
[1351867173] Processing by RecipesController#main as HTML
Logs are still shuffled, but now we can normalize, order them.
sort -f -s -k1.1,1.11 production.log | sed 's/^.............//' > sorted_production.log
(Sorter by first symbols (by timestamp) and remove timestamp by sed)
Now logs are easy to analyze.
In addition there is fix on related issue https://github.com/rails/rails/pull/7317 in rails 3.2.9
So keep this in mind.
Sorry for bad English... )

What is the difference in "Processing by Contoller#method as */*" and "Processing by BillsController#show as HTML"

All of my rails 3.2.2 ActiveRecord methods are being executed twice. I noticed that each execution is being processed differently, see the examples I grabbed from the console below...
Started GET "/api/bills/Jeremy%20Fox" for 127.0.0.1 at 2012-03-20 23:16:43 -0400
Processing by BillsController#show as HTML
Parameters: {"username"=>"Jeremy Fox"}
BillsForUsers Load (2.4ms) SELECT "bills_for_users".* FROM "bills_for_users" WHERE "bills_for_users"."billusername" = 'Jeremy Fox'
Completed 200 OK in 47ms (Views: 11.2ms | ActiveRecord: 2.4ms)
Started GET "/api/bills/Jeremy%20Fox" for 127.0.0.1 at 2012-03-20 23:16:44 -0400
Processing by BillsController#show as */*
Parameters: {"username"=>"Jeremy Fox"}
BillsForUsers Load (1.1ms) SELECT "bills_for_users".* FROM "bills_for_users" WHERE "bills_for_users"."billusername" = 'Jeremy Fox'
Completed 200 OK in 33ms (Views: 28.1ms | ActiveRecord: 1.1ms)
Can anyone explain to me why all of my ActiveRecord methods are being executed twice and/or what the difference is between Processing by BillsController#show as HTML and Processing by BillsController#show as */*?
Thanks.
It turns out the problem was actually the JSONView Chrome extension. As it states in the options menu...
Use safe method to parse HTTP response (*)
(*) : safe method forces the browser to send an extra HTTP request to get the raw HTTP content.
After spending days trying to figure out what I was doing wrong in my code, it was actually just chrome!
Hope no one else runs into this stupid issue.
-Jeremy
I've been grappling with this same issue. The HTML Validator Chrome extension is also guilty (with none of the fine print.)
In my case, I'm calling a ModestModel-backed search request, so the first hit (and rendering) was succeeding, followed by a phantom 500 as my non-DB search model was out of scope and nil on the second request.
Thanks, Jeremy!

Rails 3 not writing 500 status to log

I'm running a Rails 3.0.7 app with nginx and Passenger. I have a custom 500 page that is properly displayed when the app encounters an 500 internal error, however the actual '500' status is not being output to the logs.
I'd like to be able to periodically grep the logs to find 500 errors, but I can't seem to figure out why the actual status is not being rendered. I've even looked through the Rails code, and everything looks fine. All other status codes are successfully logged.
Here is an error-free 200 response:
Completed 200 OK in 1265ms (Views: 1262.4ms | ActiveRecord: 69.6ms | Sphinx: 0.0ms)
Here is a 500 response:
Completed in 500ms
It appears that something is supposed to be there, but is not, so spaces are output instead.
Looks like this has been resolved in Rails master, but is not in the gem for 3.0.7 yet.
https://github.com/rails/rails/commit/7927fc2ff77543a0ab151ac1cb3d60318e2dfa68

objects that are not created show on the page, and redirect fails

I tried out rails 3, and just started a new project. The following is all I typed.
rails new todo
cd todo
bundle install
rails generate scaffold Task done:boolean task:text created:date
rake db:migrate
rails server&
firefox 0.0.0:3000/tasks&
On page 0.0.0:3000/tasks/new, I filled in some values for the fields, and clicked the button to create a new task. It redirects to 0.0.0:3000/tasks with a blank page. When I manually reload the page, it shows up eight tasks even though there is supposed to be only one.
When I further click either show, edit, or destroy, it says, for e.g.: ActiveRecord::RecordNotFound in TasksController#show Couldn't find Task with ID=1.
When I reload to 0.0.0:3000/tasks, all eight tasks are still there.
What is wrong with this? Is rails corrupted on my computer?
Log
When I click 'create tasks', the terminal displays
Started GET "/tasks/new" for 127.0.0.1
at 2011-05-13 22:04:26 -0400
Processing by TasksController#new as
HTML Rendered tasks/_form.html.erb
(6.7ms) Rendered tasks/new.html.erb
within layouts/application (25.3ms)
Completed 200 OK in 35ms (Views:
27.9ms | ActiveRecord: 0.0ms)
folowed by something like this repeated eight times with x in tasks/x varying from 1 to 8:
Started POST "/tasks" for 127.0.0.1 at
2011-05-13 22:04:32 -0400 Processing
by TasksController#create as HTML
Parameters: {"utf8"=>"✓",
"authenticity_token"=>"QZHWWyE5KcZhLrzRhB4Fgnl9HGiQqNkn17O4CUfUUJU=",
"task"=>{"done"=>"0",
"task"=>"test\r\n",
"created(1i)"=>"2011",
"created(2i)"=>"5",
"created(3i)"=>"14"},
"commit"=>"Create Task"} AREL
(0.2ms) INSERT INTO "tasks" ("done",
"task", "created", "created_at",
"updated_at") VALUES ('f', 'test ',
'2011-05-14', '2011-05-14
02:04:32.065805', '2011-05-14
02:04:32.065805') Redirected to
http://0.0.0:3000/tasks/2 Completed
302 Found in 17ms [2011-05-13
22:04:32] ERROR URI::InvalidURIError:
the scheme http does not accept
registry part: 0.0.0:3000 (or bad
hostname?)
/usr/local/lib/ruby/1.9.1/uri/generic.rb:746:in
rescue in merge'
/usr/local/lib/ruby/1.9.1/uri/generic.rb:743:in
merge'
/usr/local/lib/ruby/1.9.1/webrick/httpresponse.rb:163:in
setup_header'
/usr/local/lib/ruby/1.9.1/webrick/httpresponse.rb:101:in
send_response'
/usr/local/lib/ruby/1.9.1/webrick/httpserver.rb:86:in
run'
/usr/local/lib/ruby/1.9.1/webrick/server.rb:183:in
block in start_thread'
You're trying to access your site using 0.0.0:3000 which is an invalid address (I'm actually surprised this even gives you access to the site at all).
Open 0.0.0.0:3000 in Firefox instead, and it will work perfectly!
(You can also use localhost:3000 or 127.0.0.1:3000)
Task is a reserved word in Rails. I guess thats the reason why rails is behaving in a starnge way.
For a list of other reserved words please refer to any of these links
http://cheat.errtheblog.com/s/rails_reserved_words/
http://www.yup.com/articles/2007/01/31/no-reservations-about-keywords-in-ruby-on-rails
http://oldwiki.rubyonrails.org/rails/pages/ReservedWords

Resources