Killed by RPCError

```
#2018-10-11 02:30:51 UTC agent listen INFO agent-hub The client encountered an error:
#2018-10-11 02:30:52 UTC agent listen INFO ws::io Accepted a new tcp connection from 52.78.215.10:39808.
#2018-10-11 02:30:52 UTC agent-107 INFO agent-hub Agent-107 started
#2018-10-11 02:31:14 UTC ERROR iron::iron Error handling:
Request {
url: Url { generic_url: "http://:5012/log/validator1" }
method: Get
remote_addr: V4(:57714)
local_addr: V4(0.0.0.0:5012)
}
Error was: RPCError Internal error about jsonrpc call : Timeout timed out waiting on channel
Killed
```

No source file, test, or entry point is named; the report only shows an agent-hub log with an RPCError timeout while handling a request to /log/validator1. Start by locating the RPCError and timeout handling around the agent-hub request, then reproduce the failure described by the timestamped log. Done means identifying the timeout cause and preventing the agent from being killed.

Written by the indexing model from the issue text.