Evolution always finds a way

[earthgecko]

And so do users ....:)

A continuation of https://github.com/etsy/skyline/commit/22ae09da716267a65835472da89ac31cc5cc5192#commitcomment-16606603

@yutongche - That sounds like either Horizon was not started or redis was not available on the settings.py `REDIS_SOCKET_PATH = '/tmp/redis.sock'` socket.

It is difficult to say which as Horizon is feed here via UDP, fire and forget.

So ensure that Horizon is running.  You can do that by running the following from your command line with telnet:

telnet 127.0.0.1 2025

#Trying 127.0.0.1...
#Connected to 127.0.0.1.
#Escape character is '^]'.
#^]
#telnet> quit
#Connection closed.


That will test that you can connect to Horizon on the TCP pickle port anyway (as long as you did not change the default PICKLE_PORT in settings.py).
If Horizon is running on the TCP port, let assume for now that the UDP port will be fine as well.

Secondly ensure that redis running and is available on the socket at `REDIS_SOCKET_PATH = '/tmp/redis.sock'`.  Testing a socket is possible but if we assume that you have a default redis install it will be on port 6379.

But first, ensure that the redis config has unxisocket enabled (e.g. /etc/redis/redis.conf) with:

unixsocket /tmp/redis.sock
unixsocketperm 777


If socket was not enabled and you have enabled it in the config, restart Redis.

Test redis is available via telnet:

telnet 127.0.0.1 6379


This does not test to socket, but let us assume if the sockect is enabled and redis is running, socket is working.

If Horizon and redis are available, try run seed_data.py and we shall hope that you see...

"Congratulations! The data made it in. The Horizon pipeline seems to be working."

(Note to self - add validation and detailed debug output to every stage of seed_data.py)