Configuring FIWARE Lab Node
This section provides details about how to make the proper changes in the configuration of your node in order to join FIWARE Lab. Those changes are basically related to the proper configuration of flavours and quotas and more important, related to the common way to define the available networks in a FIWARE Lab node.
Configure Flavors and Quotas
The default flavors should be:
ID | Name | Memory (MB) | Disk (Gb) | Ephemeral | Swap | vCPUs | RXTX Factor | Public |
---|---|---|---|---|---|---|---|---|
1 | m1.tiny | 512 | 1 | 0 | 1 | 1.0 | True | |
2 | m1.small | 2048 | 20 | 0 | 1 | 1.0 | True | |
3 | m1.medium | 4096 | 40 | 0 | 2 | 1.0 | True | |
4 | m1.large | 8192 | 80 | 0 | 4 | 1.0 | True |
For the nova service, the default quotas values that should be defined are the following:
Default defined quotas
Quota | Limit |
---|---|
Instances | 2 |
Cores | 4 |
RAM | 4096 |
Floating IPs | 1 |
Fixed IPs | -1 |
Metadata Items | 1024 |
Injected files | 5 |
Injected file content (bytes) | 20240 |
Injected file path (bytes) | 255 |
Key pairs | 10 |
Security Groups | 10 |
Security Group Rules | 20 |
The neutron default quotas should be:
Default defined neutron quotas
Field | Value |
---|---|
Floating IP | 1 |
Network | 5 |
Port | 20 |
Router | 1 |
Security Group | -1 |
Security Group Rule | -1 |
Subnet | 5 |
Configure OpenStack Networks
FIWARE Lab has defined a predefined name of networks to be used by all the nodes. It helps to the different services deployed on top of OpenStack to work with the correct network without any special configuration on it.
-
public-ext-net-01. This is the Public External network, a non-shared network providing a floating IP pool (i.e. subnet) that provides public, routable IPv4 addresses. Additionally, nodes can configure IPv6 dual-stack on this network in order to provide IPv6 addresses. This network is not visible to attach directly OpenStack Instances on it. It is only visible to allocate public IPs to be used by tenants.
-
node-int-net-01. A shared tenant network providing DHCP IPv4 (and IPv6 in the future) addresses. This network is visible for all tenants and therefore anyone can attach OpenStack instances on it. Any node could choose its own network range since this should not collide with other node’s networks.
There is no limitation in the use of networks and every node can configure additional networks in its OpenStack configuration. If we test this information with the CLI tool we obtain the following result if we execute the following command:
$ neutron net-list
Or using the more recent version of the CLI, the following command:
$ openstack network list
The output of networks and subnets should be:
Example returned values of: openstack network list
Id | Name | subnets |
---|---|---|
3dccc622-7200-40be-b523-0f73674db0e7 | public-ext-net-01 | 44c356e1-53ad-43ce-b3b7-816bbd1d9529 130.206.82.0/22 |
b99da016-cb02-4556-8d5f-2ce27a9a861d | node-int-net-01 | a250c7a4-4d23-4c9a-85be-3e9b367a00a1 172.16.0.0/20 |
And if we check the sub-network that we have associated to this network, through the following commands
$ neutron subnet-list
or
$ openstack subnet list
we will see something like this for the second network:
Example returned values of: openstack subnet list
Id | Name | CIDR | Allocation pools |
---|---|---|---|
a250c7a4-4d23-4c9a-85be-3e9b367a00a1 | node-int-subnet-01 | 172.16.0.0/20 | {"start": "172.16.0.2", "end": "172.16.15.254"} |