astron: Fix segfault when handling `CLIENTAGENT_ADD_INTEREST` and Interest Operation times out

[maxrdz]

The astron cluster crashes with a segmentation fault when it receives a CLIENTAGENT_ADD_INTEREST message from the AI process. Weird behavior to note: If you send this message with parent ID 0 and zone ID 0, it does not crash the astron daemon. And I have no idea why.

[maxrdz]

The temporary but very not logical fix for this issue is to comment out all the lines in the InterestOperation::timeout() function @ src/clientagent/Client.cpp.

void InterestOperation::timeout()
{
    //lock_guard<recursive_mutex> lock(m_client->m_client_lock);
    //m_client->m_log->warning() << "Interest operation timed out; forcing.\n";
    //finish(true);
}

Note: the astrond coredump points the segfault to happen when the mutex is locked, but if you remove only that line it says it happens somewhere in the InterestOperation::timeout() function, but if you remove the m_log->warning() line it will say the segfault happens in InterestOperation::finish(bool timeout)???????