Hung-Wei Chiu | 77c969e | 2020-10-23 18:13:07 +0000 | [diff] [blame] | 1 | .. |
| 2 | SPDX-FileCopyrightText: © 2020 Open Networking Foundation <support@opennetworking.org> |
| 3 | SPDX-License-Identifier: Apache-2.0 |
| 4 | |
Charles Chan | 4a10722 | 2020-10-30 17:23:48 -0700 | [diff] [blame] | 5 | Hardware Installation |
| 6 | ===================== |
Hung-Wei Chiu | 77c969e | 2020-10-23 18:13:07 +0000 | [diff] [blame] | 7 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 8 | Once the hardware has been ordered, the installation can be planned and |
| 9 | implemented. This document describes the installation of the servers and |
| 10 | software. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 11 | |
| 12 | Installation of the fabric switch hardware is covered in :ref:`OS Installation |
| 13 | - Switches <switch-install>`. |
| 14 | |
| 15 | Installation of the radio hardware is covered in :ref:`eNB Installation |
| 16 | <enb-installation>`. |
| 17 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 18 | Site Bookkeeping |
| 19 | ---------------- |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 20 | |
| 21 | The following items need to be added to `NetBox |
| 22 | <https://netbox.readthedocs.io/en/stable>`_ to describe each edge site: |
| 23 | |
| 24 | 1. Add a Site for the edge (if one doesn't already exist), which has the |
| 25 | physical location and contact information for the edge. |
| 26 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 27 | 2. Add equipment Racks to the Site (if they don't already exist). |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 28 | |
| 29 | 3. Add a Tenant for the edge (who owns/manages it), assigned to the ``Pronto`` |
| 30 | or ``Aether`` Tenant Group. |
| 31 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 32 | 4. Add a VRF (Routing Table) for the edge site. This is usually just the name |
| 33 | of the site. Make sure that ``Enforce unique space`` is checked, so that IP |
| 34 | addresses within the VRF are forced to be unique, and that the Tenant Group |
| 35 | and Tenant are set. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 36 | |
| 37 | 5. Add a VLAN Group to the edge site, which groups the site's VLANs and |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 38 | requires that they have a unique VLAN number. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 39 | |
| 40 | 6. Add VLANs for the edge site. These should be assigned a VLAN Group, the |
| 41 | Site, and Tenant. |
| 42 | |
| 43 | There can be multiple of the same VLAN in NetBox (VLANs are layer 2, and |
| 44 | local to the site), but not within the VLAN group. |
| 45 | |
| 46 | The minimal list of VLANs: |
| 47 | |
| 48 | * ADMIN 1 |
| 49 | * UPLINK 10 |
| 50 | * MGMT 800 |
| 51 | * FAB 801 |
| 52 | |
| 53 | If you have multiple deployments at a site using the same management server, |
| 54 | add additional VLANs incremented by 10 for the MGMT/FAB - for example: |
| 55 | |
| 56 | * DEVMGMT 810 |
| 57 | * DEVFAB 801 |
| 58 | |
| 59 | 7. Add IP Prefixes for the site. This should have the Tenant and VRF assigned. |
| 60 | |
| 61 | All edge IP prefixes fit into a ``/22`` sized block. |
| 62 | |
| 63 | The description of the Prefix contains the DNS suffix for all Devices that |
| 64 | have IP addresses within this Prefix. The full DNS names are generated by |
| 65 | combining the first ``<devname>`` component of the Device names with this |
| 66 | suffix. |
| 67 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 68 | An examples using the ``10.0.0.0/22`` block. There are 4 edge |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 69 | prefixes, with the following purposes: |
| 70 | |
| 71 | * ``10.0.0.0/25`` |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 72 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 73 | * Has the Server BMC/LOM and Management Switch |
| 74 | * Assign the ADMIN 1 VLAN |
| 75 | * Set the description to ``admin.<deployment>.<site>.aetherproject.net`` (or |
| 76 | ``prontoproject.net``). |
| 77 | |
| 78 | * ``10.0.0.128/25`` |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 79 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 80 | * Has the Server Management plane, Fabric Switch Management/BMC |
| 81 | * Assign MGMT 800 VLAN |
| 82 | * Set the description to ``<deployment>.<site>.aetherproject.net`` (or |
| 83 | ``prontoproject.net``). |
| 84 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 85 | * ``10.0.1.0/25`` |
| 86 | |
| 87 | * IP addresses of the qsfp0 port of the Compute Nodes to Fabric switches, devices |
| 88 | connected to the Fabric like the eNB |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 89 | * Assign FAB 801 VLAN |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 90 | * Set the description to ``fab1.<deployment>.<site>.aetherproject.net`` (or |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 91 | ``prontoproject.net``). |
| 92 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 93 | * ``10.0.1.128/25`` |
| 94 | |
| 95 | * IP addresses of the qsfp1 port of the Compute Nodes to fabric switches |
| 96 | * Assign FAB 801 VLAN |
| 97 | * Set the description to ``fab2.<deployment>.<site>.aetherproject.net`` (or |
| 98 | ``prontoproject.net``). |
| 99 | |
Zack Williams | 5fd7a23 | 2020-12-03 12:45:56 -0700 | [diff] [blame] | 100 | There also needs to be a parent range of the two fabric ranges added: |
| 101 | |
| 102 | * ``10.0.1.0/24`` |
| 103 | |
| 104 | * This is used to configure the correct routes, DNS, and TFTP servers |
| 105 | provided by DHCP to the equipment that is connected to the fabric |
| 106 | leaf switch that the management server (which provides those |
| 107 | services) is not connected to. |
| 108 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 109 | Additionally, these edge prefixes are used for Kubernetes but don't need to |
| 110 | be created in NetBox: |
| 111 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 112 | * ``10.0.2.0/24`` |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 113 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 114 | * Kubernetes Pod IP's |
| 115 | |
| 116 | * ``10.0.3.0/24`` |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 117 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 118 | * Kubernetes Cluster IP's |
| 119 | |
| 120 | 8. Add Devices to the site, for each piece of equipment. These are named with a |
| 121 | scheme similar to the DNS names used for the pod, given in this format:: |
| 122 | |
| 123 | <devname>.<deployment>.<site> |
| 124 | |
| 125 | Examples:: |
| 126 | |
| 127 | mgmtserver1.ops1.tucson |
| 128 | node1.stage1.menlo |
| 129 | |
| 130 | Note that these names are transformed into DNS names using the Prefixes, and |
| 131 | may have additional components - ``admin`` or ``fabric`` may be added after |
| 132 | the ``<devname>`` for devices on those networks. |
| 133 | |
| 134 | Set the following fields when creating a device: |
| 135 | |
| 136 | * Site |
| 137 | * Tenant |
| 138 | * Rack & Rack Position |
| 139 | * Serial number |
| 140 | |
| 141 | If a specific Device Type doesn't exist for the device, it must be created, |
| 142 | which is detailed in the NetBox documentation, or ask the OPs team for help. |
| 143 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 144 | See `Rackmount of Equipment`_ below for guidance on how equipment should be |
| 145 | mounted in the Rack. |
| 146 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 147 | 9. Add Services to the management server: |
| 148 | |
| 149 | * name: ``dns`` |
| 150 | protocol: UDP |
| 151 | port: 53 |
| 152 | |
| 153 | * name: ``tftp`` |
| 154 | protocol: UDP |
| 155 | port: 69 |
| 156 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 157 | These are used by the DHCP and DNS config to know which servers offer |
| 158 | DNS or TFTP service. |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 159 | |
| 160 | 10. Set the MAC address for the physical interfaces on the device. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 161 | |
| 162 | You may also need to add physical network interfaces if aren't already |
| 163 | created by the Device Type. An example would be if additional add-in |
| 164 | network cards were installed. |
| 165 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 166 | 11. Add any virtual interfaces to the Devices. When creating a virtual |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 167 | interface, it should have it's ``label`` field set to the physical network |
| 168 | interface that it is assigned |
| 169 | |
| 170 | These are needed are two cases for the Pronto deployment: |
| 171 | |
| 172 | 1. On the Management Server, there should bet (at least) two VLAN |
| 173 | interfaces created attached to the ``eno2`` network port, which |
| 174 | are used to provide connectivity to the management plane and fabric. |
| 175 | These should be named ``<name of vlan><vlan ID>``, so the MGMT 800 VLAN |
| 176 | would become a virtual interface named ``mgmt800``, with the label |
| 177 | ``eno2``. |
| 178 | |
| 179 | 2. On the Fabric switches, the ``eth0`` port is shared between the OpenBMC |
| 180 | interface and the ONIE/ONL installation. Add a ``bmc`` virtual |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 181 | interface with a label of ``eth0`` on each fabric switch, and check the |
| 182 | ``OOB Management`` checkbox. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 183 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 184 | 12. Create IP addresses for the physical and virtual interfaces. These should |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 185 | have the Tenant and VRF set. |
| 186 | |
| 187 | The Management Server should always have the first IP address in each |
| 188 | range, and they should be incremental, in this order. Examples are given as |
| 189 | if there was a single instance of each device - adding additional devices |
| 190 | would increment the later IP addresses. |
| 191 | |
| 192 | * Management Server |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 193 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 194 | * ``eno1`` - site provided public IP address, or blank if DHCP |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 195 | provided |
| 196 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 197 | * ``eno2`` - 10.0.0.1/25 (first of ADMIN) - set as primary IP |
| 198 | * ``bmc`` - 10.0.0.2/25 (next of ADMIN) |
| 199 | * ``mgmt800`` - 10.0.0.129/25 (first of MGMT) |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 200 | * ``fab801`` - 10.0.1.1/25 (first of FAB) |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 201 | |
| 202 | * Management Switch |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 203 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 204 | * ``gbe1`` - 10.0.0.3/25 (next of ADMIN) - set as primary IP |
| 205 | |
| 206 | * Fabric Switch |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 207 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 208 | * ``eth0`` - 10.0.0.130/25 (next of MGMT), set as primary IP |
| 209 | * ``bmc`` - 10.0.0.131/25 |
| 210 | |
| 211 | * Compute Server |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 212 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 213 | * ``eth0`` - 10.0.0.132/25 (next of MGMT), set as primary IP |
| 214 | * ``bmc`` - 10.0.0.4/25 (next of ADMIN) |
| 215 | * ``qsfp0`` - 10.0.1.2/25 (next of FAB) |
| 216 | * ``qsfp1`` - 10.0.1.3/25 |
| 217 | |
| 218 | * Other Fabric devices (eNB, etc.) |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 219 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 220 | * ``eth0`` or other primary interface - 10.0.1.4/25 (next of FAB) |
| 221 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 222 | 13. Add DHCP ranges to the IP Prefixes for IP's that aren't reserved. These are |
| 223 | done like any other IP Address, but with the ``Status`` field is set to |
| 224 | ``DHCP``, and they'll consume the entire range of IP addresses given in the |
| 225 | CIDR mask. |
| 226 | |
| 227 | For example ``10.0.0.32/27`` as a DHCP block would take up 1/4 of the ADMIN |
| 228 | prefix. |
| 229 | |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 230 | 14. Add router IP reservations to the IP Prefix for both Fabric prefixes. These |
Zack Williams | 9d94b4f | 2020-12-14 11:25:29 -0700 | [diff] [blame] | 231 | are IP addresses used by ONOS to route traffic to the other leaf, and have |
| 232 | the following attributes: |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 233 | |
Zack Williams | 5fd7a23 | 2020-12-03 12:45:56 -0700 | [diff] [blame] | 234 | - Have the last usable address in range (in the ``/25`` fabric examples |
| 235 | above, these would be ``10.0.1.126/25`` and ``10.0.1.254/25``) |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 236 | |
| 237 | - Have a ``Status`` of ``Reserved``, and the VRF, Tenant Group, and Tenant |
| 238 | set. |
| 239 | |
| 240 | - The Description must start with the word ``router``, such as: ``router |
Zack Williams | 5fd7a23 | 2020-12-03 12:45:56 -0700 | [diff] [blame] | 241 | for leaf1 Fabric`` |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 242 | |
Zack Williams | 9d94b4f | 2020-12-14 11:25:29 -0700 | [diff] [blame] | 243 | - A custom field named ``RFC3442 Routes`` is set to the CIDR IP address of |
| 244 | the opposite leaf - if the leaf's prefix is ``10.0.1.0/25`` and the |
| 245 | router IP is ``10.0.1.126/25`` then ``RFC3442 Routes`` should be set to |
| 246 | ``10.0.1.128\25`` (and the reverse - on ``10.0.1.254/25`` the ``RFC3442 |
| 247 | Routes`` would be set to be ``10.0.1.0/25``). This creates an `RFC3442 |
| 248 | Classless Static Route Option <https://tools.ietf.org/html/rfc3442>`_ |
| 249 | for the subnet in DHCP. |
| 250 | |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 251 | 15. Add Cables between physical interfaces on the devices |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 252 | |
Zack Williams | e8c3b2c | 2021-02-01 12:47:28 -0700 | [diff] [blame^] | 253 | The topology needs to match the logical diagram presented in the |
| 254 | :ref:`network_cable_plan`. Note that many of the management interfaces |
| 255 | need to be located either on the MGMT or ADMIN VLANs, and the management |
| 256 | switch is |
| 257 | used to provide that separation. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 258 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 259 | Rackmount of Equipment |
| 260 | ---------------------- |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 261 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 262 | Most of the Pronto equipment has a 19" rackmount form factor. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 263 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 264 | Guidelines for mounting this equipment: |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 265 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 266 | - The EdgeCore Wedge Switches have a front-to-back (aka "port-to-power") fan |
| 267 | configuration, so hot air exhaust is out the back of the switch near the |
| 268 | power inlets, away from the 32 QSFP network ports on the front of the switch. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 269 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 270 | - The full-depth 1U and 2U Supermicro servers also have front-to-back airflow |
| 271 | but have most of their ports on the rear of the device. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 272 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 273 | - Airflow through the rack should be in one direction to avoid heat being |
| 274 | pulled from one device into another. This means that to connect the QSFP |
| 275 | network ports from the servers to the switches, cabling should be routed |
| 276 | through the rack from front (switch) to back (server). Empty rack spaces |
| 277 | should be reserved for this purpose. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 278 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 279 | - The short-depth management HP Switch and 1U Supermicro servers should be |
| 280 | mounted on the rear of the rack. They both don't generate an appreciable |
| 281 | amount of heat, so the airflow direction isn't a significant factor in |
| 282 | racking them. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 283 | |
| 284 | Inventory |
| 285 | --------- |
| 286 | |
| 287 | Once equipment arrives, any device needs to be recorded in inventory if it: |
| 288 | |
| 289 | 1. Connects to the network (has a MAC address) |
| 290 | 2. Has a serial number |
| 291 | 3. Isn't a subcomponent (disk, add-in card, linecard, etc.) of a larger device. |
| 292 | |
| 293 | The following information should be recorded for every device: |
| 294 | |
| 295 | - Manufacturer |
| 296 | - Model |
| 297 | - Serial Number |
| 298 | - MAC address (for the primary and any management/BMC/IPMI interfaces) |
| 299 | |
| 300 | This information should be be added to the corresponding Devices ONF NetBox |
| 301 | instance. The accuracy of this information is very important as it is used in |
| 302 | bootstrapping the systems. |
| 303 | |
| 304 | Once inventory has been completed, let the Infra team know, and the pxeboot |
| 305 | configuration will be generated to have the OS preseed files corresponding to the |
| 306 | new servers based on their serial numbers. |
| 307 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 308 | Management Switch Bootstrap |
| 309 | --------------------------- |
| 310 | |
Zack Williams | e8c3b2c | 2021-02-01 12:47:28 -0700 | [diff] [blame^] | 311 | The current Pronto deployment uses an HP/Aruba 2540 24G PoE+ 4SFP+ JL356A |
| 312 | switch to run the management network and other VLAN's that are used internally. |
| 313 | |
| 314 | By default the switch will pull an IP address via DHCP and ``http://<switch IP>`` |
| 315 | will display a management webpage for the switch. You need to be able to access |
| 316 | this webpage before you can update the configuration. |
| 317 | |
| 318 | Loading the Management Switch Configuration |
| 319 | """"""""""""""""""""""""""""""""""""""""""" |
| 320 | |
| 321 | 1. Obtain a copy of the Management switch configuration file (this ends in ``.pcc``). |
| 322 | |
| 323 | 2. Open the switch web interface at ``http://<switch IP>``. You may be |
| 324 | prompted to login - the default credentials are both ``admin``: |
| 325 | |
| 326 | .. image:: images/pswi-000.png |
| 327 | :alt: User Login for switch |
| 328 | :scale: 50% |
| 329 | |
| 330 | 3. Go to the "Management" section at bottom left: |
| 331 | |
| 332 | .. image:: images/pswi-001.png |
| 333 | :alt: Update upload |
| 334 | :scale: 50% |
| 335 | |
| 336 | In the "Update" section at left, drag the configuration file into the upload |
| 337 | area, or click Browse and select it. |
| 338 | |
| 339 | 4. In the "Select switch configuration file to update" section, select |
| 340 | "config1", so it overwrites the default configuration. |
| 341 | |
| 342 | 5. In the "Select switch configuration file to update" section, select |
| 343 | "config1", so it overwrites the default configuration. Click "Update". |
| 344 | You'll be prompted to reboot the switch, which you can do with the power |
| 345 | symbol at top right. You may be prompted to select an image used to reboot - |
| 346 | the "Previously Selected" is the correct one to use: |
| 347 | |
| 348 | .. image:: images/pswi-003.png |
| 349 | :alt: Switch Image Select |
| 350 | :scale: 30% |
| 351 | |
| 352 | 6. Wait for the switch to reboot: |
| 353 | |
| 354 | .. image:: images/pswi-004.png |
| 355 | :alt: Switch Reboot |
| 356 | :scale: 50% |
| 357 | |
| 358 | The switch is now configured with the correct VLANs for Pronto Use. If you |
| 359 | go to Interfaces > VLANs should see a list of VLANs configured on the |
| 360 | switch: |
| 361 | |
| 362 | .. image:: images/pswi-005.png |
| 363 | :alt: Mgmt VLANs |
| 364 | :scale: 50% |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 365 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 366 | Software Bootstrap |
| 367 | ------------------ |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 368 | |
| 369 | Management Server Bootstrap |
| 370 | """"""""""""""""""""""""""" |
| 371 | |
| 372 | The management server is bootstrapped into a customized version of the standard |
| 373 | Ubuntu 18.04 OS installer. |
| 374 | |
| 375 | The `iPXE boot firmware <https://ipxe.org/>`_. is used to start this process |
| 376 | and is built using the steps detailed in the `ipxe-build |
| 377 | <https://gerrit.opencord.org/plugins/gitiles/ipxe-build>`_. repo, which |
| 378 | generates both USB and PXE chainloadable boot images. |
| 379 | |
| 380 | Once a system has been started using these images started, these images will |
| 381 | download a customized script from an external webserver to continue the boot |
| 382 | process. This iPXE to webserver connection is secured with mutual TLS |
| 383 | authentication, enforced by the nginx webserver. |
| 384 | |
| 385 | The iPXE scripts are created by the `pxeboot |
| 386 | <https://gerrit.opencord.org/plugins/gitiles/ansible/role/pxeboot>`_ role, |
| 387 | which creates both a boot menu, downloads the appropriate binaries for |
| 388 | bootstrapping an OS installation, and creates per-node installation preseed files. |
| 389 | |
| 390 | The preseed files contain configuration steps to install the OS from the |
| 391 | upstream Ubuntu repos, as well as customization of packages and creating the |
| 392 | ``onfadmin`` user. |
| 393 | |
Zack Williams | e8c3b2c | 2021-02-01 12:47:28 -0700 | [diff] [blame^] | 394 | Creating a bootable USB drive |
| 395 | ''''''''''''''''''''''''''''' |
| 396 | |
| 397 | 1. Get a USB key. Can be tiny as the uncompressed image is floppy sized |
| 398 | (1.4MB). Download the USB image file (``<date>_onf_ipxe.usb.zip``) on the |
| 399 | system you're using to write the USB key, and unzip it. |
| 400 | |
| 401 | 2. Put a USB key in the system you're using to create the USB key, then |
| 402 | determine which USB device file it's at in ``/dev``. You might look at the |
| 403 | end of the ``dmesg`` output on Linux/Unix or the output of ``diskutil |
| 404 | list`` on macOS. |
| 405 | |
| 406 | Be very careful about this, as if you accidentally overwrite some other disk in |
| 407 | your system that would be highly problematic. |
| 408 | |
| 409 | 3. Write the image to the device:: |
| 410 | |
| 411 | $ dd if=/path/to/20201116_onf_ipxe.usb of=/dev/sdg |
| 412 | 2752+0 records in |
| 413 | 2752+0 records out |
| 414 | 1409024 bytes (1.4 MB, 1.3 MiB) copied, 2.0272 s, 695 kB/s |
| 415 | |
| 416 | You may need to use `sudo` for this. |
| 417 | |
| 418 | Boot and Image Management Server |
| 419 | '''''''''''''''''''''''''''''''' |
| 420 | |
| 421 | 1. Connect a USB keyboard and VGA monitor to the management node. Put the USB |
| 422 | Key in one of the management node's USB ports (port 2 or 3): |
| 423 | |
| 424 | .. image:: images/mgmtsrv-000.png |
| 425 | :alt: Management Server Ports |
| 426 | :scale: 50% |
| 427 | |
| 428 | 2. Turn on the management node, and press the F11 key as it starts to get into |
| 429 | the Boot Menu: |
| 430 | |
| 431 | .. image:: images/mgmtsrv-001.png |
| 432 | :alt: Management Server Boot Menu |
| 433 | :scale: 50% |
| 434 | |
| 435 | 3. Select the USB key (in this case "PNY USB 2.0", your options may vary) and press return. You should see iPXE load: |
| 436 | |
| 437 | .. image:: images/mgmtsrv-002.png |
| 438 | :alt: iPXE load |
| 439 | :scale: 50% |
| 440 | |
| 441 | 4. A menu will appear which displays the system information and DHCP discovered |
| 442 | network settings (your network must provide the IP address to the management |
| 443 | server via DHCP): |
| 444 | |
| 445 | Use the arrow keys to select "Ubuntu 18.04 Installer (fully automatic)": |
| 446 | |
| 447 | .. image:: images/mgmtsrv-003.png |
| 448 | :alt: iPXE Menu |
| 449 | :scale: 50% |
| 450 | |
| 451 | There is a 10 second default timeout if left untouched (it will continue the |
| 452 | system boot process) so restart the system if you miss the 10 second window. |
| 453 | |
| 454 | 5. The Ubuntu 18.04 installer will be downloaded and booted: |
| 455 | |
| 456 | .. image:: images/mgmtsrv-004.png |
| 457 | :alt: Ubuntu Boot |
| 458 | :scale: 50% |
| 459 | |
| 460 | 6. Then the installer starts and takes around 10 minutes to install (depends on |
| 461 | your connection speed): |
| 462 | |
| 463 | .. image:: images/mgmtsrv-005.png |
| 464 | :alt: Ubuntu Install |
| 465 | :scale: 50% |
| 466 | |
| 467 | |
| 468 | 7. At the end of the install, the system will restart and present you with a |
| 469 | login prompt: |
| 470 | |
| 471 | .. image:: images/mgmtsrv-006.png |
| 472 | :alt: Ubuntu Install Complete |
| 473 | :scale: 50% |
| 474 | |
| 475 | |
| 476 | Management Server Configuration |
| 477 | ''''''''''''''''''''''''''''''' |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 478 | |
| 479 | Once the OS is installed on the management server, Ansible is used to remotely |
| 480 | install software on the management server. |
| 481 | |
| 482 | To checkout the ONF ansible repo and enter the virtualenv with the tooling:: |
| 483 | |
| 484 | mkdir infra |
| 485 | cd infra |
| 486 | repo init -u ssh://<your gerrit username>@gerrit.opencord.org:29418/infra-manifest |
| 487 | repo sync |
| 488 | cd ansible |
| 489 | make galaxy |
| 490 | source venv_onfansible/bin/activate |
| 491 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 492 | Obtain the ``undionly.kpxe`` iPXE artifact for bootstrapping the compute |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 493 | servers, and put it in the ``playbook/files`` directory. |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 494 | |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 495 | Next, create an inventory file to access the NetBox API. An example is given |
| 496 | in ``inventory/example-netbox.yml`` - duplicate this file and modify it. Fill |
| 497 | in the ``api_endpoint`` address and ``token`` with an API key you get out of |
| 498 | the NetBox instance. List the IP Prefixes used by the site in the |
| 499 | ``ip_prefixes`` list. |
| 500 | |
| 501 | Next, run the ``scripts/netbox_edgeconfig.py`` to generate a host_vars file for |
| 502 | the management server. Assuming that the management server in the edge is |
| 503 | named ``mgmtserver1.stage1.menlo``, you'd run:: |
| 504 | |
| 505 | python scripts/netbox_edgeconfig.py inventory/my-netbox.yml > inventory/host_vars/mgmtserver1.stage1.menlo.yml |
| 506 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 507 | One manual change needs to be made to this output - edit the |
| 508 | ``inventory/host_vars/mgmtserver1.stage1.menlo.yml`` file and add the following |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 509 | to the bottom of the file, replacing the IP addresses with the management |
| 510 | server IP address for each segment. |
| 511 | |
| 512 | In the case of the Fabric that has two leaves and IP ranges, add the Management |
| 513 | server IP address used for the leaf that it is connected to, and then add a |
| 514 | route for the other IP address range for the non-Management-connected leaf that |
Zack Williams | 5fd7a23 | 2020-12-03 12:45:56 -0700 | [diff] [blame] | 515 | is via the Fabric router address in the connected leaf range. |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 516 | |
| 517 | This configures the `netplan <https://netplan.io>`_ on the management server, |
Zack Williams | 5fd7a23 | 2020-12-03 12:45:56 -0700 | [diff] [blame] | 518 | and creates a SNAT rule for the UE range route, and will be automated away |
| 519 | soon:: |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 520 | |
| 521 | # added manually |
| 522 | netprep_netplan: |
| 523 | ethernets: |
| 524 | eno2: |
| 525 | addresses: |
| 526 | - 10.0.0.1/25 |
| 527 | vlans: |
| 528 | mgmt800: |
| 529 | id: 800 |
| 530 | link: eno2 |
| 531 | addresses: |
| 532 | - 10.0.0.129/25 |
| 533 | fabr801: |
| 534 | id: 801 |
| 535 | link: eno2 |
| 536 | addresses: |
Zack Williams | e8cb121 | 2020-12-03 09:48:58 -0700 | [diff] [blame] | 537 | - 10.0.1.129/25 |
| 538 | routes: |
| 539 | - to: 10.0.1.0/25 |
| 540 | via: 10.0.1.254 |
Zack Williams | 5fd7a23 | 2020-12-03 12:45:56 -0700 | [diff] [blame] | 541 | metric: 100 |
| 542 | |
| 543 | netprep_nftables_nat_postrouting: > |
| 544 | ip saddr 10.0.1.0/25 ip daddr 10.168.0.0/20 counter snat to 10.0.1.129; |
| 545 | |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 546 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 547 | Using the ``inventory/example-aether.ini`` as a template, create an |
| 548 | :doc:`ansible inventory <ansible:user_guide/intro_inventory>` file for the |
| 549 | site. Change the device names, IP addresses, and ``onfadmin`` password to match |
| 550 | the ones for this site. The management server's configuration is in the |
| 551 | ``[aethermgmt]`` and corresponding ``[aethermgmt:vars]`` section. |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 552 | |
| 553 | Then, to configure a management server, run:: |
| 554 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 555 | ansible-playbook -i inventory/sitename.ini playbooks/aethermgmt-playbook.yml |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 556 | |
| 557 | This installs software with the following functionality: |
| 558 | |
| 559 | - VLANs on second Ethernet port to provide connectivity to the rest of the pod. |
| 560 | - Firewall with NAT for routing traffic |
| 561 | - DHCP and TFTP for bootstrapping servers and switches |
| 562 | - DNS for host naming and identification |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 563 | - HTTP server for serving files used for bootstrapping switches |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 564 | - Downloads the Tofino switch image |
| 565 | - Creates user accounts for administrative access |
Zack Williams | 34c30e5 | 2020-11-16 10:55:00 -0700 | [diff] [blame] | 566 | |
| 567 | Compute Server Bootstrap |
| 568 | """""""""""""""""""""""" |
| 569 | |
| 570 | Once the management server has finished installation, it will be set to offer |
| 571 | the same iPXE bootstrap file to the computer. |
| 572 | |
| 573 | Each node will be booted, and when iPXE loads select the ``Ubuntu 18.04 |
| 574 | Installer (fully automatic)`` option. |
Zack Williams | a7c170f | 2020-11-25 12:59:49 -0700 | [diff] [blame] | 575 | |
| 576 | The nodes can be controlled remotely via their BMC management interfaces - if |
| 577 | the BMC is at ``10.0.0.3`` a remote user can SSH into them with:: |
| 578 | |
| 579 | ssh -L 2443:10.0.0.3:443 onfadmin@<mgmt server ip> |
| 580 | |
| 581 | And then use their web browser to access the BMC at:: |
| 582 | |
| 583 | https://localhost:2443 |
| 584 | |
| 585 | The default BMC credentials for the Pronto nodes are:: |
| 586 | |
| 587 | login: ADMIN |
| 588 | password: Admin123 |
| 589 | |
Zack Williams | 9026f53 | 2020-11-30 11:34:32 -0700 | [diff] [blame] | 590 | The BMC will also list all of the MAC addresses for the network interfaces |
| 591 | (including BMC) that are built into the logic board of the system. Add-in |
| 592 | network cards like the 40GbE ones used in compute servers aren't listed. |
| 593 | |
| 594 | To prepare the compute nodes, software must be installed on them. As they |
| 595 | can't be accessed directly from your local system, a :ref:`jump host |
| 596 | <ansible:use_ssh_jump_hosts>` configuration is added, so the SSH connection |
| 597 | goes through the management server to the compute systems behind it. Doing this |
| 598 | requires a few steps: |
| 599 | |
| 600 | First, configure SSH to use Agent forwarding - create or edit your |
| 601 | ``~/.ssh/config`` file and add the following lines:: |
| 602 | |
| 603 | Host <management server IP> |
| 604 | ForwardAgent yes |
| 605 | |
| 606 | Then try to login to the management server, then the compute node:: |
| 607 | |
| 608 | $ ssh onfadmin@<management server IP> |
| 609 | Welcome to Ubuntu 18.04.5 LTS (GNU/Linux 5.4.0-54-generic x86_64) |
| 610 | ... |
| 611 | onfadmin@mgmtserver1:~$ ssh onfadmin@10.0.0.138 |
| 612 | Welcome to Ubuntu 18.04.5 LTS (GNU/Linux 5.4.0-54-generic x86_64) |
| 613 | ... |
| 614 | onfadmin@node2:~$ |
| 615 | |
| 616 | Being able to login to the compute nodes from the management node means that |
| 617 | SSH Agent forwarding is working correctly. |
| 618 | |
| 619 | Verify that your inventory (Created earlier from the |
| 620 | ``inventory/example-aether.ini`` file) includes an ``[aethercompute]`` section |
| 621 | that has all the names and IP addresses of the compute nodes in it. |
| 622 | |
| 623 | Then run a ping test:: |
| 624 | |
| 625 | ansible -i inventory/sitename.ini -m ping aethercompute |
| 626 | |
| 627 | It may ask you about authorized keys - answer ``yes`` for each host to trust the keys:: |
| 628 | |
| 629 | The authenticity of host '10.0.0.138 (<no hostip for proxy command>)' can't be established. |
| 630 | ECDSA key fingerprint is SHA256:... |
| 631 | Are you sure you want to continue connecting (yes/no/[fingerprint])? yes |
| 632 | |
| 633 | You should then see a success message for each host:: |
| 634 | |
| 635 | node1.stage1.menlo | SUCCESS => { |
| 636 | "changed": false, |
| 637 | "ping": "pong" |
| 638 | } |
| 639 | node2.stage1.menlo | SUCCESS => { |
| 640 | "changed": false, |
| 641 | "ping": "pong" |
| 642 | } |
| 643 | ... |
| 644 | |
| 645 | Once you've seen this, run the playbook to install the prerequisites (Terraform |
| 646 | user, Docker):: |
| 647 | |
| 648 | ansible-playbook -i inventory/sitename.ini playbooks/aethercompute-playbook.yml |
| 649 | |
| 650 | Note that Docker is quite large and may take a few minutes for installation |
| 651 | depending on internet connectivity. |
| 652 | |
| 653 | Now that these compute nodes have been brought up, the rest of the installation |
| 654 | can continue. |