Ask Your Question

sendu's profile - activity

2019-07-25 01:13:43 -0500 received badge  Famous Question (source)
2019-07-25 01:13:43 -0500 received badge  Notable Question (source)
2017-11-09 08:23:32 -0500 asked a question How do you get server fault details from nova? ( shows that a server can have a "faults" section in the JSON response. But how do you get to see that?

I've tried getting from http://[ip:port]/v2.1/[...]/servers/362cfec0-ecf3-4f95-a565-1669682bce06

And I get:

{"server": {"status": "ERROR", [...]

But that does not contain any faults section.

2017-01-27 06:45:54 -0500 received badge  Scholar (source)
2017-01-27 06:45:50 -0500 received badge  Popular Question (source)
2017-01-27 05:37:28 -0500 commented answer How does ports quota get used up?

Thanks. Do I take it from this that regardless of anything else, a new instance will always use up exactly 1 port?

2017-01-27 05:36:08 -0500 received badge  Supporter (source)
2017-01-27 03:40:58 -0500 asked a question How does ports quota get used up?

I have a user where neutron quota-show indicates a ports quota of 50.

They are able to bring up 42 servers that belong to a security group that has 3 tcp ports open.

On attempting to bring up the 43rd server they get an error:

Maximum number of ports exceeded

(They might also have additional servers up on a different network.)

I take it from this that the "ports" quota has nothing to do with tcp ports in the security group?

How can I calculate ahead of time how much of their ports quota a new server would use, to avoid requesting a server that will exceed the quota?

2016-09-28 13:03:08 -0500 asked a question create instance from an instance?

Let's say a user uses some unknown method of creating an instance (nova compute node), and they have ssh'd to it. (Presumably they must have created an ssh key and network etc. for this to work)

Now on this openstack instance they will run my software. I want my software to launch new instances that my software can ssh to. No user or external access to these instances is required.

What's the best way for my software to achieve this?

  1. Do I need the user to set up all the usual authentication environment variables manually?
  2. Is there any way to know that I'm running on an openstack node?
  3. How do I find out the server id of the instance I'm running on, and the network it is attached to?
  4. Will I have to create my own keypair for use with the instances I will create?

I'm using the gophercloud API in Golang, but any advice on the proper way to go about this would be appreciated.