Skip to content
This repository has been archived by the owner on Apr 6, 2021. It is now read-only.

Server that uses deepstream client has CPU 100% when deepstream server stopped #104

Open
jamesalexscott opened this issue Aug 28, 2017 · 3 comments
Labels

Comments

@jamesalexscott
Copy link

jamesalexscott commented Aug 28, 2017

I'm running a Jersey/Spring based API on Tomcat 8 that uses the deepstream Java client to connect to deepstream and then send various RPC requests. I've been developing and testing on Windows and everything has been fine.

Now I'm moving it into production on AWS (latest Linux on a t2.small). If I stop the deepstream server, the CPU immediately goes to 100% for the Tomcat process. When I restart deepstream and the API logs back in to deepstream but the CPU still remains at 100%. Only a Tomcat restart fixes it.

I'm also using the client in Android and it has the same issue. Though CPU goes from 5% to 60%

@honorcode
Copy link

honorcode commented Aug 28, 2017

We just saw the same issue on our java clients when restarting aws ds servers.
Even with 'wrapped' re-connection logic (full client.close(), client=null, client=new DeepstreamClient, client.login) - no recovery, our service nodes need full restarts.

@jamesalexscott
Copy link
Author

It wasn't the change from Windows to Linux. It was the change from ws to wss. Seems there may be an infinite loop issue with the underlying Java-Websocket.

Doesn't look like it will be fixed until 1.3.5

TooTallNate/Java-WebSocket#424

@WolframHempel
Copy link
Member

WolframHempel commented Aug 29, 2017 via email

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
Projects
None yet
Development

No branches or pull requests

4 participants