openstack/tripleo-docs
Code Issues Proposed changes

Merge "Clarify DCN requirements for provider networks"

Browse Source
This commit is contained in:
Zuul 2019-04-25 15:28:53 +00:00 committed by Gerrit Code Review
commit d5adb8342e
1 changed files with 40 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
doc/source/install/advanced_deployment/distributed_compute_node.rst
View File

@ -51,7 +51,10 @@ Loss of an edge site
^^^^^^^^^^^^^^^^^^^^
Running Nova VM instances will keep running. If stopped running, you need the
control plane back to recover the stopped or crashed workloads.
control plane back to recover the stopped or crashed workloads. If Neutron DHCP
agent is centralized, and we are forwarding DHCP requests to the central site,
any VMs that are trying to renew their IPs will eventually time out and lose
connectivity.
.. note:: A single Compute service failure normally affects only its edge site
without additional downtime induced for neighbor edge sites or the central
@ -76,8 +79,9 @@ Network (OVN). There is also BGPVPN and its backend specific choices.
Network recommendations
^^^^^^^^^^^^^^^^^^^^^^^
Traditional provider networks with backbone routing at the edge may fulfill or
complement a custom distributed routing solution, like L3 Spine-Leaf topology.
Traditional or external provider networks with backbone routing at the edge may
fulfill or complement a custom distributed routing solution, like L3 Spine-Leaf
topology.
.. note:: Neutron SDN backends that involve tunnelling may be sub-optimal for
Edge DCN cases because of the known issues 1808594_ and 1808062_.
@ -85,6 +89,23 @@ complement a custom distributed routing solution, like L3 Spine-Leaf topology.
.. _1808594: https://bugs.launchpad.net/tripleo/+bug/1808594
.. _1808062: https://bugs.launchpad.net/tripleo/+bug/1808062
For dynamic IPv4 and stateful IPv6 IPAM cases, you will also need DHCP on those
provider networks in order to assign IPs to VM instances. External provider
networks usually require no Neutron DHCP agents and handle IPAM (and
routing) on its own. While for traditional or
`Routed Provider Networks <https://docs.openstack.org/neutron/latest/admin/config-routed-networks.html>`_,
when there is no L2 connectivity to edge over WAN, and Neutron DHCP agents are
placed on controllers at the central site, you should have a DHCP relay on
every provider network. Alternatively, DHCP agents need to be moved to the edge.
Such setups also require highly reliable links between remote and central sites.
.. note:: Neither of DHCP relays/agents at compute nodes, nor routed/external
provider networks are tested or automated via TripleO Heat Templates. You would
have to have those configured manually for your DCN environments.
.. note:: OVN leverages DVR and does not require running Neutron DHCP/L3 agents,
which might as well simplify particular DCN setups.
That said, when there is a network failure that disconnects the edge off the
central site, there is no SLA for recovery time but only what the provider
networks or a particular SDN choice can guarantee. For switched/routed/MPLS
@ -95,11 +116,22 @@ back on various standards that are relevant here.
Config-drive/cloud-init details
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
The simplest solution we recommend for DCN would involve only provider networks
at the edge. For that case, it is also recommended to use config-drive or
another configuration mechanism other than cloud-init. Otherwise, the latter
requires a `169.254.169.254/32` route for the provider routers to forward data
to the metadata service.
Config-drive uses virtual media capabilities of the BMC controller, so that no
DHCP is required for VMs to obtain IP addresses at edge sites. This is
the most straightforward solution. This does require that the WAN between the
remote site and central site is live during deployment of a VM, but after that
the VM can run independently without a connection to the central site.
.. note:: Config-drive may be a tricky for VMs that do not support
cloud-init, like some appliance VMs. It may be that such ones (or other VMs
that do not support config-drive) will have to be configured with a static IP
that matches the Neutron port.
The simplest solution we recommend for DCN would involve only external provider
networks at the edge. For that case, it is also recommended to use either
config-drive, or IPv6 SLAAC, or another configuration mechanism other than
those requiring a `169.254.169.254/32` route for the provider routers to forward
data to the metadata service.
IPv6 details
^^^^^^^^^^^^