New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
CachingClientConnectionFactory NPE [INT-3652] #7612
Comments
Sergey Shcherbakov commented There looks to be another scenario where the fix in PR would not be sufficient. Nevertheless, since the singleUse is always "on" on the TcpNetConnections comping out from the wrapped TCP connection factory, the reading loop starts here: Since there is always a listener and singleUse is set to true. an results in lots of ERROR level messages in the logs. |
Gary Russell commented Yes; I think it's time to bite the bullet and get rid of this funky "self-closing" connection stuff (have the adapter(s) do the close). I'll open another JIRA issue. |
Gary Russell commented Although that's unrelated to your unnecessary logs issue - I think we can simply not register the cache as a listener if the cache itself doesn't have a listener. |
Gary Russell commented Reopening to remove the listener altogether. |
Gary Russell commented After further review; the correct does indeed require #7614 because the cache relies on the self-closing behavior to return the connection by being aware that the interceptor wrapped the connection. |
Gary Russell commented 4.1.3 and 4.0.7 are now available. |
Gary Russell opened INT-3652 and commented
When a CCCF is used with no listener and an exception occurs on a socket, an error message is sent to the onMessage method, which expects there to be a listener.
The reader thread will get an NPE.
This was a regression caused by the addition of the
ErrorMessage
handling.Affects: 4.0.6, 4.1.2
Referenced from: pull request #1380
Backported to: 4.0.7
1 votes, 2 watchers
The text was updated successfully, but these errors were encountered: