Fixed OOD shell timeout issue by injecting heartbeat function

[wtripp180901]

For some reason the OOD shell doesn't seem to be affected by either ClientAliveInterval in the sshd config or ServerAliveInterval in the ssh config. This only happens via the OOD shell, I tried ssh-ing from the login pod to itself as OOD would and the issue doesn't occur. This could potentially be a bug
https://discourse.openondemand.org/t/ssh-keep-alive-shell-access-on-ondemand/384/14
OSC/ondemand#2269
https://discourse.openondemand.org/t/novnc-and-shell-session-timeout-after-1-minute/1622

I found a horrible fix here:
https://discourse.openondemand.org/t/novnc-and-shell-session-timeout-after-1-minute/1622/6

I'm mounting the file from the OOD source with those changes in before startup

diff of modified app.js file with the original:

101,102c101,102
<   let match = url.match(host_path_rx),
<   hostname = null,
---
>   let match = url.match(host_path_rx),
>   hostname = null,
112,117d111
< function noop() {}
<
< function heartbeat() {
<   this.isAlive = true;
< }
<
160,162d153
<   ws.isAlive = true;
<   ws.on('pong', heartbeat);
<
168,175d158
<
< const interval = setInterval(function ping() {
<   wss.clients.forEach(function each(ws) {
<     if (ws.isAlive === false) return ws.terminate();
<     ws.isAlive = false;
<     ws.ping(noop);
<   });
< }, 30000);

[sjpb]

@wtripp180901 could you find a way of providing a diff between the upstream file and the one this PR injects please? It can just go into the first comment above.