Handshake_timeout on RabbitMQ using python and pika from remote vm

Question

I have several vm's around the world, all identical and with the same python code (And same library version of pika 0.9.13). Only one of them is failing because of handshake_timeout:

=ERROR REPORT==== 17-Jun-2015::12:27:07 ===
closing AMQP connection <0.290.0> (YY.YY.YY.YY:47232 -> XX.XX.XX.XX:5672):
{handshake_timeout,handshake}

that is taken from the rabbitMQ log, in the client code I see something like

ERROR:pika.adapters.base_connection:Connection to XX.XX.XX.XX:5672 failed: timeout
Unhandled exception in thread started by <function listen_for_start_download_message at 0x1e5bcf8>

I have tested the connection with telnet following this guide: http://rubybunny.info/articles/troubleshooting.html

I already increased the handshake_timeout to 40000 milliseconds and the ssl_handshake_timeout to 20000 milliseconds and the error persists, the ping to the rabbitmq server from that machine is a little bit higher than in the other machines but nothing abnormal (136ms). Has anyone found a similar problem sometime or can anyone recommend another way to test the rabbitMQ connection?

Answer 1

I'm really not sure if this could be the solution to anyone with this problem but in my case the problem was resolved by installing these libraries on the problematic vm's: amqp-tools and librabbitmq1 simply using apt-get install. I have no idea why the other vms had those libraries and the problematic ones did not but that was the case.