Bad gateway for some requests
Sometimes I got bad gateway for app akqbridgenew.apphb.com
Here are log examples -
2018-08-23T21:30:43.000+00:00 appharbor nginx 93.85.55.71 - "GET /signalr/poll?transport=longPolling&connectionToken=ygWcxUAk%2FWoHMBHtlXcXt9O9tEWUPl%2FggAGXXB7j6H5jJUlSyte86XSvAW34vAkhnCQZtThm05W9iAO7ijatJonzOZHYg0Fyo5B4BpUNGdSvCQ%2BK3p6U6zp8eXQMNnDawGDODOE%2BeydxRFvQXZmLBg%3D%3D&groupsToken=TJ9BQ698DeQSrkuP%2BIJLByhUmJPFiYik%2B5z%2FLhzf%2BUUw3Q6uQ7e2ZS65AfiJ6Y%2Flhsf5vQ%2F1Z4%2BpdI5zDeWIT%2BHS5OgGttUOVPYfo7OEbnVMgSZc%2FLRMWjYT92qooBdgs3E1RZo4kG6fUGVRoLeR5d0M%2BhGqvunnFNpG9BeycbI%3D&messageId=d-7C61F24-B%2C0%7CC%2C0%7CD%2C0%7CE%2C2%7CF%2C0%7CG%2C2&connectionData=%5B%7B%22name%22%3A%22bridgehub%22%7D%5D&tid=0&callback=jQuery203027971648081849754_1535059697196&_=1535059697201 HTTP/1.1" 502 568 "https://akqbridge.com/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/68.0.3440.106 Safari/537.36" "source=nginx measure#http.response_time=6.382s measure#http.bytes.sent=718"
2018-08-23T21:30:43.000+00:00 appharbor nginx 93.85.55.71 - "POST /game/play-card HTTP/1.1" 502 568 "https://akqbridge.com/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/68.0.3440.106 Safari/537.36" "source=nginx measure#http.response_time=4.621s measure#http.bytes.sent=718"
Discussions are closed to public comments.
If you need help with AppHarbor please
start a new discussion.
Keyboard shortcuts
Generic
? | Show this help |
---|---|
ESC | Blurs the current field |
Comment Form
r | Focus the comment reply box |
---|---|
^ + ↩ | Submit the comment |
You can use Command ⌘
instead of Control ^
on Mac
Support Staff 1 Posted by rune on 24 Aug, 2018 08:23 PM
Hi,
This can sometimes happen if the web worker process terminates unexpectedly (which may be tricky to debug), but is often caused by exceptions that occur outside the request context (e.g. in a "background"/async thread). Do you perform any such processing on a regular basis? Also, the issue appears to be quite intermittent, and I'm curious if this started happening recently and could be related to recent changes to your application's code?
Best,
Rune
2 Posted by antonputau on 26 Aug, 2018 07:35 PM
akqbridge is a gaming site. This case (502) is related to concrete game for concrete user - looks like a bug in application. My problem is that locally app with all prod settings has no errors. Only on appharbor I am catching this error and I can reproduce this case. Are there any server logs related directly to IIS, not to nginx?
3 Posted by antonputau on 31 Aug, 2018 10:21 PM
I can confirm I have no errors on serverside. I have no error entries catched by elmah (http module that tracks unhandled errors) when 502 occurs.
Support Staff 4 Posted by rune on 02 Sep, 2018 04:39 PM
Hi,
Sorry about the slow response -- I noticed that more errors that might very well be causing/related to the issues you're experiencing here are being logged in the application's Errors page. Did you perhaps disable some logic in your app that might've handled the exceptions?
Also, if you're able to reproduce it locally I suspect it might also be more straightforward to fix the issue? Of course depends on what type of issue we're dealing with here but I'm curious if you have more details on that?
Best,
Rune