Tutorial problems on a VM

I'm trying out GF by following the docs [1] and I found a few issue when running the tutorial in vFabric_GemFire_701_b40235/SampleCode/tutorial:

The  tutorial on src/com/gemstone/gemfire/tutorial/storage/GemfireDAO.java defines the locator as "localhost[port]". I changed that to "ip[port]" (both for peer and clients) and recompiled. The tutorial still runs fine, but as soon as I try to start a client in a different node (a virtual machine) it stops working:

1. When I run the peer I can see it connecting to the locator (with netstat on the first host), but it takes a long time to connect (over 10 secs) and the data is not shared between peers (should it be shared on this case?).
2. When I run the client it doesn't even start, it throws a NoSubscriptionServersAvailableException after printing "Could not initialize a primary queue on startup. No queue servers available". I haven't found much info about this exception, what does it mean?

Should the tutorial work on a different node for both setups (client and peer)? I'm not sure about the peer setup but I understand it should work for the client.

Thanks

PS: There's no firewall or anything else limiting the connection between VMs.

EDIT:

I also found the following:

1. Start the servers and the locator on $host1
2. Start gfsh on $host2
3. Run"connect --locator=$host1[$port]". Gfsh will print something like this:
     Connecting to Locator at [host=$host1, port=$port] ..
     Connecting to Manager at [host=$host1, port=1099] ..
     Connection refused to host: 127.0.1.1; nested exception is: ...

Is that exception related to the manager connection? If so, why would the host2 try to connect to localhost? (Of course $host1 != 127.0.1.1)

[1] vFabric Documentation Center