-
-
Notifications
You must be signed in to change notification settings - Fork 134
Error Domain=SRWebSocketErrorDomain Code=2132 "received bad response code from server 404" #33
Comments
My best guess is that you don't have the proper URL for your LiveQuery server. Can you show us the URL you used? |
sure, we are using nginx 1.8.1 distributing using distelli of the parse sample server with the live query on and with mount path as / and with 'npm ws' called before 'npm start'. Parse node server is running fine and accessible via dashboard. Any quick way to test if the web socket is working? upstream parseNodeServer {
|
Everything there seems good... Other than using the existing LiveQuery clients we have, not that I know of. What about the URL you configured the client with? Is it possible you're missing a trailing slash (that happens to be important for the default node implementation of LiveQuery)? |
on the iOS client side tried it with the trailing slash Since the live query piggy back the parser server URL, if the parse server URL is working, we assume the live query URL should be good. |
@yz, since you are using This wasn't yet released, but is available on the latest master. |
@nlutsenko @richardjrossiii found the bug related to this issue, it was due to that the path component in the Client.swift is hardcoded. It should be appending since our server, i suspect others, are using a reverse proxy with a location route. So the current implementation will fail for any parser server with a location route. Change the following line in the Client.swift fixed this 404 error. That said, after this 404 issue is resolved, was able to run and connect to the Chat room example, however, couple of new issues, messages are not showing up in the Dashboard (due to schema?) but did get created in the database and updates via mlab admin/browser did not trigger the subscription. That said, is there a working sample for iOS client app? |
Yup, that would be it, we are discussing right now on the best way to resolve this that would allow setting the path. We are currently requiring for the path to be hardcoded into the URL according to the setup of Live Query Server here. In regards to messages not showing up in the Dashboard - are you using Parse.com dashboard? Be aware that changes through mlab admin should not trigger the subscription, because the Live Query Server is no aware of them. Meaning that changes are triggered only if they come to the server itself and not directly to the database. |
@yz the dashboard and SDKs use a batch endpoint that requires your mount path to be the same as the URL you provide when you initialize your Parse SDK. I highly recommend you avoid any URL rewriting in your reverse proxy. |
@nlutsenko yes, using the Parse's dashboard before and it does not show the messages. Switched to use local dashboard does show the messages. However, after connecting to the Room, it does not print out the existing messages, and by modifying via local dashboard (which goes thru the server) still does not trigger the live query. |
@drew-gross our server has few node apps, and we are using nginx and each one has its own location, we are not rewriting url, simply route the location to each node.js app and without any any mount path simply / This happen to be necessary for the server push notification to work as raised in one of the thread (Non root routes break Push Notifications #87) |
This hasn't been necessary for a long time with LiveQuery, and should allow this to operate properly behind a reverse proxy or other scenarios where precise path matters. Fixes #33.
This hasn't been necessary for a long time with LiveQuery, and should allow this to operate properly behind a reverse proxy or other scenarios where precise path matters. Fixes #33.
I'm having issue while deploying ParseLiveQuery on AWS. It's working locally but when I deploy on AWS and try to test. I get this error. |
we are also facing the same issue for WebSocket connection in ios swift app |
same issue here |
We are getting this error with live query. Anyone has any tips on what might caused it?
The text was updated successfully, but these errors were encountered: