CONFIRMED HANG Redis disconnection causes a hang and timeout

[vidia]

Testing methodology: Using code on redis-errs I ran the app and started a siege. Then I took down the redis instance during siege. App froze. Restarting the redis instance seems to fix this.

This doesnt explain what causes the disconnect in prod, but it is a confirmed hang. We do not fall back to the wash alert website when redis disconnects, unlike what the get_all.js code seems to suggest.

[vidia]

I believe that changes on redis-errs fix most of this. If you take on this issue, take a look at that.