fix: Flush wavefront output sender on error to clean up broken connections

[mikolajpe]

Required for all PRs:

• Updated associated README.md.
• Wrote appropriate unit tests.
• Pull request title or commits are in conventional commit format

resolves #8812

1. update wavefront-sdk-go to the most recent version
2. added flushing of the sender once it encounters an error, without it what's happening is that the connection is never cleaned and the client is constantly trying to send messages via broken connection. This means if the wavefront-proxy goes down and the connection breaks (for example when container has crushed or you are simply deploying new version of it) telegraf will never again acquire new connection and resume sending, with this fix it will

[telegraf-tiger]

Thanks so much for the pull request!
🤝 ✒️ Just a reminder that the CLA has not yet been signed, and we'll need it before merging. Please sign the CLA when you get a chance, then post a comment here saying !signed-cla

[mikolajpe]

!signed-cla

[telegraf-tiger]

Thanks so much for the pull request!
🤝 ✒️ Just a reminder that the CLA has not yet been signed, and we'll need it before merging. Please sign the CLA when you get a chance, then post a comment here saying !signed-cla

[srebhan]

Thanks for fixing this issue @mikolajprzybysz! One comment about which error to return. What do you think?

[srebhan]

To be honest I would prefer

Suggested change

	w.Log.Errorf("wavefront sending error: %v", err)
	return w.sender.Flush()
	if flushErr := w.sender.Flush(); flushErr != nil {
	w.Log.Errorf("wavefront flushing error: %v", flushErr)
	}
	return err

as the sending error is the relevant one to return IMO. What do you think?

[srebhan]

Looks good to me. Thanks @mikolajprzybysz for the quick fix!