After upgrading to v2.0+ I started getting this error on every deploy. The node that it points to is always a MQTT node. Sometimes an IN node and sometimes an OUT node. Also, the deploy itself takes about 10-15 seconds now. I did a search on the topic and it talked about Config nodes. That issue does not seem to apply here.
Are you running the broker in Node-RED or externally?
There was an issue reported along these lines when using one of the contrib nodes that host an mqtt broker in Node-RED - although that dates back to the 2.0 release.
I am using an internal node. It is the Aedes broker node. I am running NR v 2.1.2
Generally, in my opinion, it is better to use Mosquitto. It is very easy to install and configure.
1 Like
I may need to do that. The reason that I went with a broker node was it made it easier to manage the flow via making it a project and backing it up with github. When deploying to other devices I did not need to also install mosquitto or another non-node broker. I was just being lazy. 
1 Like
@knolleary should I move to a non-node based broker?
That's up to you. I'm not aware of a fix being forthcoming for the close time out issue. You could either follow up with the maintainer of the nodes you're using, or switch to an external broker.
Personally, I share @Colin's view that it's better to use Mosquitto outside of Node-RED.
Will do. Thanks
Hi, I am the maintainer of node-red-contrib-aedes.
Just to let you know the issue does not happen if you deploy Modified Flows or Modified Nodes only
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.
Out of interest, has a workaround been developed e.g. a flow that can be run to disconnect the connection between client and broker, before doing a complete "Restart Flows" and, if required, another flow to reconnect after restart?
Given the time it's taking me, I would prefer to click on an inject node (for example) before restarting my flows and then click on another inject node (for example) to reconnect afterwards.
Appreciate it may not be a big deal for most, but it affects me when I'm doing a lot of work in NR developing / testing.
Switch to using mosquitto and the problem should go away.
Alex has opened another thread on switching to mosquitto 
I reopened this one to carry on the discussion about using Aedes
I would prefer to get Aedes "fixed" but in the meantime I have switched to mosquitto.
For those following this thread, if you have the opportunity to update to node-red 2.2.2, this problem should now be improved. The other part will require an update to the aedes broker (reported in the repo in issue #54
