pronto_deployment_guide/hw_installation.rst

..
   SPDX-FileCopyrightText: © 2020 Open Networking Foundation <support@opennetworking.org>
   SPDX-License-Identifier: Apache-2.0

Hardware Installation
=====================

Once the hardware has been ordered, the installation can be planned and
implemented. This document describes the installation of the servers and
software.

Installation of the fabric switch hardware is covered in :ref:`OS Installation
- Switches <switch-install>`.

Installation of the radio hardware is covered in :ref:`eNB Installation
<enb-installation>`.

Site Bookkeeping
----------------

The following items need to be added to `NetBox
<https://netbox.readthedocs.io/en/stable>`_ to describe each edge site:

1. Add a Site for the edge (if one doesn't already exist), which has the
   physical location and contact information for the edge.

2. Add equipment Racks to the Site (if they don't already exist).

3. Add a Tenant for the edge (who owns/manages it), assigned to the ``Pronto``
   or ``Aether`` Tenant Group.

4. Add a VRF (Routing Table) for the edge site. This is usually just the name
   of the site.  Make sure that ``Enforce unique space`` is checked, so that IP
   addresses within the VRF are forced to be unique, and that the Tenant Group
   and Tenant are set.

5. Add a VLAN Group to the edge site, which groups the site's VLANs and
   requires that they have a unique VLAN number.

6. Add VLANs for the edge site.  These should be assigned a VLAN Group, the
   Site, and Tenant.

   There can be multiple of the same VLAN in NetBox (VLANs are layer 2, and
   local to the site), but not within the VLAN group.

   The minimal list of VLANs:

     * ADMIN 1
     * UPLINK 10
     * MGMT 800
     * FAB 801

   If you have multiple deployments at a site using the same management server,
   add additional VLANs incremented by 10 for the MGMT/FAB - for example:

     * DEVMGMT 810
     * DEVFAB 801

7. Add IP Prefixes for the site. This should have the Tenant and VRF assigned.

   All edge IP prefixes fit into a ``/22`` sized block.

   The description of the Prefix contains the DNS suffix for all Devices that
   have IP addresses within this Prefix. The full DNS names are generated by
   combining the first ``<devname>`` component of the Device names with this
   suffix.

   An examples using the ``10.0.0.0/22`` block. There are 4 edge
   prefixes, with the following purposes:

     * ``10.0.0.0/25``

        * Has the Server BMC/LOM and Management Switch
        * Assign the ADMIN 1 VLAN
        * Set the description to ``admin.<deployment>.<site>.aetherproject.net`` (or
          ``prontoproject.net``).

     * ``10.0.0.128/25``

        * Has the Server Management plane, Fabric Switch Management/BMC
        * Assign MGMT 800 VLAN
        * Set the description to ``<deployment>.<site>.aetherproject.net`` (or
          ``prontoproject.net``).

     * ``10.0.1.0/25``

        * IP addresses of the qsfp0 port of the Compute Nodes to Fabric switches, devices
          connected to the Fabric like the eNB
        * Assign FAB 801 VLAN
        * Set the description to ``fab1.<deployment>.<site>.aetherproject.net`` (or
          ``prontoproject.net``).

     * ``10.0.1.128/25``

        * IP addresses of the qsfp1 port of the Compute Nodes to fabric switches
        * Assign FAB 801 VLAN
        * Set the description to ``fab2.<deployment>.<site>.aetherproject.net`` (or
          ``prontoproject.net``).

   There also needs to be a parent range of the two fabric ranges added:

     * ``10.0.1.0/24``

        * This is used to configure the correct routes, DNS, and TFTP servers
          provided by DHCP to the equipment that is connected to the fabric
          leaf switch that the management server (which provides those
          services) is not connected to.

   Additionally, these edge prefixes are used for Kubernetes but don't need to
   be created in NetBox:

     * ``10.0.2.0/24``

        * Kubernetes Pod IP's

     * ``10.0.3.0/24``

        * Kubernetes Cluster IP's

8. Add Devices to the site, for each piece of equipment. These are named with a
   scheme similar to the DNS names used for the pod, given in this format::

     <devname>.<deployment>.<site>

   Examples::

     mgmtserver1.ops1.tucson
     node1.stage1.menlo

   Note that these names are transformed into DNS names using the Prefixes, and
   may have additional components - ``admin`` or ``fabric`` may be added after
   the ``<devname>`` for devices on those networks.

   Set the following fields when creating a device:

     * Site
     * Tenant
     * Rack & Rack Position
     * Serial number

   If a specific Device Type doesn't exist for the device, it must be created,
   which is detailed in the NetBox documentation, or ask the OPs team for help.

   See `Rackmount of Equipment`_ below for guidance on how equipment should be
   mounted in the Rack.

9. Add Services to the management server:

    * name: ``dns``
      protocol: UDP
      port: 53

    * name: ``tftp``
      protocol: UDP
      port: 69

   These are used by the DHCP and DNS config to know which servers offer
   DNS or TFTP service.

10. Set the MAC address for the physical interfaces on the device.

   You may also need to add physical network interfaces if  aren't already
   created by the Device Type.  An example would be if additional add-in
   network cards were installed.

11. Add any virtual interfaces to the Devices. When creating a virtual
    interface, it should have it's ``label`` field set to the physical network
    interface that it is assigned

    These are needed are two cases for the Pronto deployment:

     1. On the Management Server, there should bet (at least) two VLAN
        interfaces created attached to the ``eno2`` network port, which
        are used to provide connectivity to the management plane and fabric.
        These should be named ``<name of vlan><vlan ID>``, so the MGMT 800 VLAN
        would become a virtual interface named ``mgmt800``, with the label
        ``eno2``.

     2. On the Fabric switches, the ``eth0`` port is shared between the OpenBMC
        interface and the ONIE/ONL installation.  Add a ``bmc`` virtual
        interface with a label of ``eth0`` on each fabric switch, and check the
        ``OOB Management`` checkbox.

12. Create IP addresses for the physical and virtual interfaces.  These should
    have the Tenant and VRF set.

    The Management Server should always have the first IP address in each
    range, and they should be incremental, in this order. Examples are given as
    if there was a single instance of each device - adding additional devices
    would increment the later IP addresses.

      * Management Server

          * ``eno1`` - site provided public IP address, or blank if DHCP
            provided

          * ``eno2`` - 10.0.0.1/25 (first of ADMIN) - set as primary IP
          * ``bmc`` - 10.0.0.2/25 (next of ADMIN)
          * ``mgmt800`` - 10.0.0.129/25 (first of MGMT)
          * ``fab801`` - 10.0.1.1/25 (first of FAB)

      * Management Switch

          * ``gbe1`` - 10.0.0.3/25 (next of ADMIN) - set as primary IP

      * Fabric Switch

          * ``eth0`` - 10.0.0.130/25 (next of MGMT), set as primary IP
          * ``bmc`` - 10.0.0.131/25

      * Compute Server

          * ``eth0`` - 10.0.0.132/25 (next of MGMT), set as primary IP
          * ``bmc`` - 10.0.0.4/25 (next of ADMIN)
          * ``qsfp0`` - 10.0.1.2/25 (next of FAB)
          * ``qsfp1`` - 10.0.1.3/25

      * Other Fabric devices (eNB, etc.)

          * ``eth0`` or other primary interface - 10.0.1.4/25 (next of FAB)

13. Add DHCP ranges to the IP Prefixes for IP's that aren't reserved. These are
    done like any other IP Address, but with the ``Status`` field is set to
    ``DHCP``, and they'll consume the entire range of IP addresses given in the
    CIDR mask.

    For example ``10.0.0.32/27`` as a DHCP block would take up 1/4 of the ADMIN
    prefix.

14. Add router IP reservations to the IP Prefix for both Fabric prefixes. These
    are IP addresses used by ONOS to route traffic to the other leaf, and have
    the following attributes:

    - Have the last usable address in range (in the ``/25`` fabric examples
      above, these would be ``10.0.1.126/25`` and ``10.0.1.254/25``)

    - Have a ``Status`` of ``Reserved``, and the VRF, Tenant Group, and Tenant
      set.

    - The Description must start with the word ``router``, such as: ``router
      for leaf1 Fabric``

    - A custom field named ``RFC3442 Routes`` is set to the CIDR IP address of
      the opposite leaf - if the leaf's prefix is ``10.0.1.0/25`` and the
      router IP is ``10.0.1.126/25`` then ``RFC3442 Routes`` should be set to
      ``10.0.1.128\25`` (and the reverse - on ``10.0.1.254/25`` the ``RFC3442
      Routes`` would be set to be ``10.0.1.0/25``).  This creates an `RFC3442
      Classless Static Route Option <https://tools.ietf.org/html/rfc3442>`_
      for the subnet in DHCP.

15. Add Cables between physical interfaces on the devices

    The topology needs to match the logical diagram presented in the
    :ref:`network_cable_plan`.  Note that many of the management interfaces
    need to be located either on the MGMT or ADMIN VLANs, and the management
    switch is
    used to provide that separation.

Rackmount of Equipment
----------------------

Most of the Pronto equipment has a 19" rackmount form factor.

Guidelines for mounting this equipment:

- The EdgeCore Wedge Switches have a front-to-back (aka "port-to-power") fan
  configuration, so hot air exhaust is out the back of the switch near the
  power inlets, away from the 32 QSFP network ports on the front of the switch.

- The full-depth 1U and 2U Supermicro servers also have front-to-back airflow
  but have most of their ports on the rear of the device.

- Airflow through the rack should be in one direction to avoid heat being
  pulled from one device into another.  This means that to connect the QSFP
  network ports from the servers to the switches, cabling should be routed
  through the rack from front (switch) to back (server).  Empty rack spaces
  should be reserved for this purpose.

- The short-depth management HP Switch and 1U Supermicro servers should be
  mounted on the rear of the rack.  They both don't generate an appreciable
  amount of heat, so the airflow direction isn't a significant factor in
  racking them.

Inventory
---------

Once equipment arrives, any device needs to be recorded in inventory if it:

1. Connects to the network (has a MAC address)
2. Has a serial number
3. Isn't a subcomponent (disk, add-in card, linecard, etc.) of a larger device.

The following information should be recorded for every device:

- Manufacturer
- Model
- Serial Number
- MAC address (for the primary and any management/BMC/IPMI interfaces)

This information should be be added to the corresponding Devices ONF NetBox
instance.  The accuracy of this information is very important as it is used in
bootstrapping the systems.

Once inventory has been completed, let the Infra team know, and the pxeboot
configuration will be generated to have the OS preseed files corresponding to the
new servers based on their serial numbers.

Management Switch Bootstrap
---------------------------

The current Pronto deployment uses an HP/Aruba 2540 24G PoE+ 4SFP+ JL356A
switch to run the management network and other VLAN's that are used internally.

By default the switch will pull an IP address via DHCP and ``http://<switch IP>``
will display a management webpage for the switch. You need to be able to access
this webpage before you can update the configuration.

Loading the Management Switch Configuration
"""""""""""""""""""""""""""""""""""""""""""

1. Obtain a copy of the Management switch configuration file (this ends in ``.pcc``).

2.  Open the switch web interface at ``http://<switch IP>``. You may be
    prompted to login - the default credentials are both ``admin``:

    .. image:: images/pswi-000.png
       :alt: User Login for switch
       :scale: 50%

3. Go to the "Management" section at bottom left:

   .. image:: images/pswi-001.png
       :alt: Update upload
       :scale: 50%

   In the "Update" section at left, drag the configuration file into the upload
   area, or click Browse and select it.

4. In the "Select switch configuration file to update" section, select
   "config1", so it overwrites the default configuration.

5. In the "Select switch configuration file to update" section, select
   "config1", so it overwrites the default configuration. Click "Update".
   You'll be prompted to reboot the switch, which you can do with the power
   symbol at top right. You may be prompted to select an image used to reboot -
   the "Previously Selected" is the correct one to use:

   .. image:: images/pswi-003.png
       :alt: Switch Image Select
       :scale: 30%

6. Wait for the switch to reboot:

   .. image:: images/pswi-004.png
       :alt: Switch Reboot
       :scale: 50%

   The switch is now configured with the correct VLANs for Pronto Use.  If you
   go to Interfaces > VLANs should see a list of VLANs configured on the
   switch:

   .. image:: images/pswi-005.png
       :alt: Mgmt VLANs
       :scale: 50%

Software Bootstrap
------------------

Management Server Bootstrap
"""""""""""""""""""""""""""

The management server is bootstrapped into a customized version of the standard
Ubuntu 18.04 OS installer.

The `iPXE boot firmware <https://ipxe.org/>`_. is used to start this process
and is built using the steps detailed in the `ipxe-build
<https://gerrit.opencord.org/plugins/gitiles/ipxe-build>`_. repo, which
generates both USB and PXE chainloadable boot images.

Once a system has been started using these images started, these images will
download a customized script from  an external webserver to continue the boot
process. This iPXE to webserver connection is secured with mutual TLS
authentication, enforced by the nginx webserver.

The iPXE scripts are created by the `pxeboot
<https://gerrit.opencord.org/plugins/gitiles/ansible/role/pxeboot>`_ role,
which creates both a boot menu, downloads the appropriate binaries for
bootstrapping an OS installation, and creates per-node installation preseed files.

The preseed files contain configuration steps to install the OS from the
upstream Ubuntu repos, as well as customization of packages and creating the
``onfadmin`` user.

Creating a bootable USB drive
'''''''''''''''''''''''''''''

1. Get a USB key. Can be tiny as the uncompressed image is floppy sized
   (1.4MB).  Download the USB image file (``<date>_onf_ipxe.usb.zip``) on the
   system you're using to write the USB key, and unzip it.

2. Put a USB key in the system you're using to create the USB key, then
   determine which USB device file it's at in ``/dev``.  You might look at the
   end of the ``dmesg`` output on Linux/Unix or the output of ``diskutil
   list`` on macOS.

   Be very careful about this, as if you accidentally overwrite some other disk in
   your system that would be highly problematic.

3. Write the image to the device::

      $ dd if=/path/to/20201116_onf_ipxe.usb of=/dev/sdg
      2752+0 records in
      2752+0 records out
      1409024 bytes (1.4 MB, 1.3 MiB) copied, 2.0272 s, 695 kB/s

  You may need to use `sudo` for this.

Boot and Image Management Server
''''''''''''''''''''''''''''''''

1. Connect a USB keyboard and VGA monitor to the management node.  Put the USB
   Key in one of the management node's USB ports (port 2 or 3):

   .. image:: images/mgmtsrv-000.png
       :alt: Management Server Ports
       :scale: 50%

2. Turn on the management node, and press the F11 key as it starts to get into
   the Boot Menu:

   .. image:: images/mgmtsrv-001.png
       :alt: Management Server Boot Menu
       :scale: 50%

3. Select the USB key (in this case "PNY USB 2.0", your options may vary) and press return. You should see iPXE load:

   .. image:: images/mgmtsrv-002.png
       :alt: iPXE load
       :scale: 50%

4. A menu will appear which displays the system information and DHCP discovered
   network settings (your network must provide the IP address to the management
   server via DHCP):

   Use the arrow keys to select "Ubuntu 18.04 Installer (fully automatic)":

   .. image:: images/mgmtsrv-003.png
       :alt: iPXE Menu
       :scale: 50%

   There is a 10 second default timeout if left untouched (it will continue the
   system boot process) so restart the system if you miss the 10 second window.

5. The Ubuntu 18.04 installer will be downloaded and booted:

   .. image:: images/mgmtsrv-004.png
       :alt: Ubuntu Boot
       :scale: 50%

6. Then the installer starts and takes around 10 minutes to install (depends on
   your connection speed):

   .. image:: images/mgmtsrv-005.png
       :alt: Ubuntu Install
       :scale: 50%


7. At the end of the install, the system will restart and present you with a
   login prompt:

   .. image:: images/mgmtsrv-006.png
       :alt: Ubuntu Install Complete
       :scale: 50%


Management Server Configuration
'''''''''''''''''''''''''''''''

Once the OS is installed on the management server, Ansible is used to remotely
install software on the management server.

To checkout the ONF ansible repo and enter the virtualenv with the tooling::

  mkdir infra
  cd infra
  repo init -u ssh://<your gerrit username>@gerrit.opencord.org:29418/infra-manifest
  repo sync
  cd ansible
  make galaxy
  source venv_onfansible/bin/activate

Obtain the ``undionly.kpxe`` iPXE artifact for bootstrapping the compute
servers, and put it in the ``playbook/files`` directory.

Next, create an inventory file to access the NetBox API.  An example is given
in ``inventory/example-netbox.yml`` - duplicate this file and modify it. Fill
in the ``api_endpoint`` address and ``token`` with an API key you get out of
the NetBox instance.  List the IP Prefixes used by the site in the
``ip_prefixes`` list.

Next, run the ``scripts/netbox_edgeconfig.py`` to generate a host_vars file for
the management server.  Assuming that the management server in the edge is
named ``mgmtserver1.stage1.menlo``, you'd run::

  python scripts/netbox_edgeconfig.py inventory/my-netbox.yml > inventory/host_vars/mgmtserver1.stage1.menlo.yml

One manual change needs to be made to this output - edit the
``inventory/host_vars/mgmtserver1.stage1.menlo.yml`` file and add the following
to the bottom of the file, replacing the IP addresses with the management
server IP address for each segment.

In the case of the Fabric that has two leaves and IP ranges, add the Management
server IP address used for the leaf that it is connected to, and then add a
route for the other IP address range for the non-Management-connected leaf that
is via the Fabric router address in the connected leaf range.

This configures the `netplan <https://netplan.io>`_ on the management server,
and creates a SNAT rule for the UE range route, and will be automated away
soon::

  # added manually
  netprep_netplan:
    ethernets:
      eno2:
        addresses:
          - 10.0.0.1/25
    vlans:
      mgmt800:
        id: 800
        link: eno2
        addresses:
          - 10.0.0.129/25
      fabr801:
        id: 801
        link: eno2
        addresses:
          - 10.0.1.129/25
        routes:
          - to: 10.0.1.0/25
            via: 10.0.1.254
            metric: 100

  netprep_nftables_nat_postrouting: >
    ip saddr 10.0.1.0/25 ip daddr 10.168.0.0/20 counter snat to 10.0.1.129;


Using the ``inventory/example-aether.ini`` as a template, create an
:doc:`ansible inventory <ansible:user_guide/intro_inventory>` file for the
site. Change the device names, IP addresses, and ``onfadmin`` password to match
the ones for this site.  The management server's configuration is in the
``[aethermgmt]`` and corresponding ``[aethermgmt:vars]`` section.

Then, to configure a management server, run::

  ansible-playbook -i inventory/sitename.ini playbooks/aethermgmt-playbook.yml

This installs software with the following functionality:

- VLANs on second Ethernet port to provide connectivity to the rest of the pod.
- Firewall with NAT for routing traffic
- DHCP and TFTP for bootstrapping servers and switches
- DNS for host naming and identification
- HTTP server for serving files used for bootstrapping switches
- Downloads the Tofino switch image
- Creates user accounts for administrative access

Compute Server Bootstrap
""""""""""""""""""""""""

Once the management server has finished installation, it will be set to offer
the same iPXE bootstrap file to the computer.

Each node will be booted, and when iPXE loads select the ``Ubuntu 18.04
Installer (fully automatic)`` option.

The nodes can be controlled remotely via their BMC management interfaces - if
the BMC is at ``10.0.0.3`` a remote user can SSH into them with::

  ssh -L 2443:10.0.0.3:443 onfadmin@<mgmt server ip>

And then use their web browser to access the BMC at::

  https://localhost:2443

The default BMC credentials for the Pronto nodes are::

  login: ADMIN
  password: Admin123

The BMC will also list all of the MAC addresses for the network interfaces
(including BMC) that are built into the logic board of the system. Add-in
network cards like the 40GbE ones used in compute servers aren't listed.

To prepare the compute nodes, software must be installed on them.  As they
can't be accessed directly from your local system, a :ref:`jump host
<ansible:use_ssh_jump_hosts>` configuration is added, so the SSH connection
goes through the management server to the compute systems behind it. Doing this
requires a few steps:

First, configure SSH to use Agent forwarding - create or edit your
``~/.ssh/config`` file and add the following lines::

  Host <management server IP>
    ForwardAgent yes

Then try to login to the management server, then the compute node::

  $ ssh onfadmin@<management server IP>
  Welcome to Ubuntu 18.04.5 LTS (GNU/Linux 5.4.0-54-generic x86_64)
  ...
  onfadmin@mgmtserver1:~$ ssh onfadmin@10.0.0.138
  Welcome to Ubuntu 18.04.5 LTS (GNU/Linux 5.4.0-54-generic x86_64)
  ...
  onfadmin@node2:~$

Being able to login to the compute nodes from the management node means that
SSH Agent forwarding is working correctly.

Verify that your inventory (Created earlier from the
``inventory/example-aether.ini`` file) includes an ``[aethercompute]`` section
that has all the names and IP addresses of the compute nodes in it.

Then run a ping test::

  ansible -i inventory/sitename.ini -m ping aethercompute

It may ask you about authorized keys - answer ``yes`` for each host to trust the keys::

  The authenticity of host '10.0.0.138 (<no hostip for proxy command>)' can't be established.
  ECDSA key fingerprint is SHA256:...
  Are you sure you want to continue connecting (yes/no/[fingerprint])? yes

You should then see a success message for each host::

  node1.stage1.menlo | SUCCESS => {
      "changed": false,
      "ping": "pong"
  }
  node2.stage1.menlo | SUCCESS => {
      "changed": false,
      "ping": "pong"
  }
  ...

Once you've seen this, run the playbook to install the prerequisites (Terraform
user, Docker)::

  ansible-playbook -i inventory/sitename.ini playbooks/aethercompute-playbook.yml

Note that Docker is quite large and may take a few minutes for installation
depending on internet connectivity.

Now that these compute nodes have been brought up, the rest of the installation
can continue.