I am getting an error while reload Socket error communicating with splunkd (error=The read operation timed out), path = /services/deployment/server/config/_reload and also its taking a lot of time to restart. But its restarting successfully.
The restart of the deployment server would reload the serverclass. No additional reload required.
You should still do a reload. I've gotten this error before and have found that it helps if you specify the class that you made adjustments to:
./splunk reload deploy-server -class newclass
I would also suggest that you open a support ticket, because I believe this is a bug.