vagrant status shows only first 2 nodes

[huliz]

In case more than 2 nodes are started, vagrant still sees only node-01 and node-02. It is not possible to ssh into higher nodes via vagrant, nor halt them. If issuing vagrant halt, the master and 2 first nodes are halted but the rest is up.

[pires]

This happens because when one uses environment variable NUM_INSTANCES to set the number of minions to spawn, it actually doesn't get persisted anywhere where Vagrantfile can read it and falls back to the default 2.

[pires]

Please do note that if you were using non default settings to startup your cluster you must also use those exact settings when invoking vagrant {up,ssh,destroy} to communicate with any of the nodes in the cluster as otherwise things may not behave as you'd expect.

So, if you ran NUM_INSTANCES=4 vagrant up you'll need to NUM_INSTANCES=4 vagrant status in order to show all the nodes. This is not optimal, but right now it's what works.

[pires]

This same issue applies for all environment variables that we set but are not reused during cluster lifecycle. We should find a way to write-and-read this info from a persistent media, i.e. a file.

[janroos]

To persist variables, I am using it like this: echo NUM_INSTANCES=1 MASTER_MEM=1024 MASTER_CPUS=2 NODE_MEM=2048 NODE_CPUS=2 vagrant "$@" > v chmod +x v ./v up

[pires]

Interesting @janroos. Thanks for sharing. But I would like something like a trigger on the Vagrantfile that would read values from a file. Such values would only be preceded by environment variables previously set.