Skip to content

Commit

Permalink
Second BPG DT
Browse files Browse the repository at this point in the history
Includes:
- BGP
- Networker nodes
- DVR
- expose-tenant-networks disabled
- ovn routing
- NBDB driver
- FRR: IP-level BGP peers
- FRR: different ASN per rack
- Tempest (neutron+octavia) + Tobiko (scenario+sanity+faults)
  • Loading branch information
eduolivares committed Jan 12, 2024
1 parent 71a1047 commit 223bcd6
Showing 1 changed file with 55 additions and 0 deletions.
55 changes: 55 additions & 0 deletions examples/dt/bgp/ovn-routing-nbdb-driver/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,55 @@
# RHOSO Deployed Topology - BGP using kernel routing and OVN NB DB driver

## General information
| Revision | Change | Date |
|--------: | :-------------------- | :--------------: |
| v0.1 | Initial publication | 2024-01-12 |

## Node topology
| Node role | bm/vm | amount |
| --------------------------------------------- | ----- | ------ |
| Openshift master/worker combo-node cluster | vm | 3 |
| Compute nodes | vm | 3 |
| Networker nodes | vm | 3 |
| Leaf routers | vm | 6 |
| Spine routers | vm | 2 |
| External routers | vm | 1 |
| External VMs | vm | 1 |

## Services, enabled features and configurations
| Service | configuration | Lock-in coverage? |
| ------------------------------------------- | ---------------------------------------------------------------------------------------------- | ------------------ |
| Neutron | ML2/OVN, DVR | Must have |
| Octavia | | Must have |
| OVN-BGP-Agent | ovn-routing, nbdb-driver expose-tenant-networks disabled, sync-interval set to a high value | Must have |
| FRR | different ASN per rack, IP-level BGP peers | Must have |

## Considerations/Constraints
1. Virtual networks should be created to connect the VMs between them.
2. All the VMs that are neither Openstack nor Openshift nodes need to be
properly configured in order to support the BGP protocol.
3. The spine/leaf topology separates the overcloud nodes into different L2
network segments, called racks. Each rack includes one compute, one
networker and two leafs.
4. A separate provisioning network is necessary to install Openstack on those
nodes.
5. Once Openstack is installed on them, controlplane and dataplane connectivity
between them and with external machines (extvm) is achieved with the BGP
protocol.
6. Tests are executed from the extvm machine, in order to verify external
connectivity.

## Testing tree
| Test framework | Stage to run | Special configuration | Test case to report |
| ------------------------ | ------------ | --------------------------------------------------------------------- | :----------------------: |
| Tempest/neutron+octavia | stage 5 | skips due to usupported features (multicast, FIP port-forwarding, etc)| scenario |
| Tobiko/Faults | stage 6 | | sanity, scenario, faults |

## Stages
All stages must be executed in the order listed below. Everything is required unless otherwise indicated.
1. [Pre-deployment: create virtual networks, virtual routers and nodes](bgp-pre-deplyment.md)
2. [Install the OpenStack K8S operators and their dependencies](../../common/)
3. [Configure networking and deploy the Openstack control plane](bgp-control-plane.md)
4. [Configure and deploy the data plane](bgp-data-plane)
5. [Execute non destructive testing](bgp-tempest.md)
6. [Execute HA testing](bgp-tobiko.md)

0 comments on commit 223bcd6

Please sign in to comment.