User Tools

Site Tools


services:networking

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
services:networking [2020/02/13 16:40]
djgalloway [Table]
services:networking [2021/06/03 03:21] (current)
djgalloway [Hardware]
Line 1: Line 1:
 ====== Networking ====== ====== Networking ======
  
-==== Networks ====+===== Networks ​=====
  
-The Sepia internal LANs are divided up into four /20 networks described below.+The Sepia internal LANs are divided up into four /20 networks ​(netmask 255.255.240.0) ​described below.
 |                     ^ front.sepia.ceph.com ​       ^ back.sepia.ceph.com ​         ^ ipmi.sepia.ceph.com ​           ^ Other                                                                                 ^ |                     ^ front.sepia.ceph.com ​       ^ back.sepia.ceph.com ​         ^ ipmi.sepia.ceph.com ​           ^ Other                                                                                 ^
 ^ Available IP Range  | 172.21.0.1 - 172.21.15.250 ​ | 172.21.16.1 - 172.21.31.250 ​ | 172.21.32.1 - 172.21.47.250 ​   | 172.21.48.1 - 172.21.63.250 ​                                                          | ^ Available IP Range  | 172.21.0.1 - 172.21.15.250 ​ | 172.21.16.1 - 172.21.31.250 ​ | 172.21.32.1 - 172.21.47.250 ​   | 172.21.48.1 - 172.21.63.250 ​                                                          |
Line 11: Line 11:
  
 In January 2020, 4 new small subnets were created for the [[hardware:​officinalis]] nodes to each have an IP on different subnets. ​ NIC1 is on VLAN104, NIC2 on VLAN105, etc.  At present, the only switch configured to use these subnets is the QFX5200 purchased specifically for the Officinalis nodes but other switches could be configured if desired. In January 2020, 4 new small subnets were created for the [[hardware:​officinalis]] nodes to each have an IP on different subnets. ​ NIC1 is on VLAN104, NIC2 on VLAN105, etc.  At present, the only switch configured to use these subnets is the QFX5200 purchased specifically for the Officinalis nodes but other switches could be configured if desired.
 +
 +==== Officinalis ====
  
 [[services:​DNS]] and [[services:​DHCP]] were configured to manage these VLANs in https://​github.com/​ceph/​ceph-sepia-secrets/​pull/​463. [[services:​DNS]] and [[services:​DHCP]] were configured to manage these VLANs in https://​github.com/​ceph/​ceph-sepia-secrets/​pull/​463.
Line 19: Line 21:
 ^ VLAN ID             | 104                          | 105                          | 106                          | 107                          | ^ VLAN ID             | 104                          | 105                          | 106                          | 107                          |
  
-==== Hardware ====+===== Hardware ​=====
 All the switches in the Sepia lab are either Juniper EX4300 1Gb or QFX5100 10Gb (SFP+) except: All the switches in the Sepia lab are either Juniper EX4300 1Gb or QFX5100 10Gb (SFP+) except:
   * A [[https://​opendcim.engineering.redhat.com/​devices.php?​DeviceID=7960|Mellanox SX1012]] private switch connected to the [[hardware:​rhoda]] and [[hardware:​incerta]] nodes   * A [[https://​opendcim.engineering.redhat.com/​devices.php?​DeviceID=7960|Mellanox SX1012]] private switch connected to the [[hardware:​rhoda]] and [[hardware:​incerta]] nodes
   * A QFX5200 in A09 for the Officinalis nodes   * A QFX5200 in A09 for the Officinalis nodes
 +  * An uplinked QFX5120 in B12 for the All Flash (name TBD) systems
 +  * A NON-uplinked [[https://​wiki.sepia.ceph.com/​doku.php?​id=hardware:​robsoni#​qfx-rgw|QFX5120]] in A15 for RGW Scale testing gear
  
 **The Sepia Networking core is === OUT === of 40G ports. ​ No more switches can be added without purchasing additional equipment.** **The Sepia Networking core is === OUT === of 40G ports. ​ No more switches can be added without purchasing additional equipment.**
  
-==== Typical Switch config ====+===== Typical Switch config ​=====
 In racks that only have 1Gb networking (e.g., [[hardware:​Mira]]),​ the top-of-rack switch is most likely split In racks that only have 1Gb networking (e.g., [[hardware:​Mira]]),​ the top-of-rack switch is most likely split
   * Ports 0-23 assigned to VLAN 100 (front)   * Ports 0-23 assigned to VLAN 100 (front)
Line 35: Line 39:
 The infra rack (houses RHEV, teuthology, gitbuilder.ceph.com),​ AA-06, has a 10Gb switch with no standard port assignments (IOW, it's kinda random). ​ I believe any port that's not already in use isn't configured. The infra rack (houses RHEV, teuthology, gitbuilder.ceph.com),​ AA-06, has a 10Gb switch with no standard port assignments (IOW, it's kinda random). ​ I believe any port that's not already in use isn't configured.
  
-==== Switch Port Config Changes ====+===== Switch Port Config Changes ​=====
  
 Red Hat IT manages all switches in the Sepia lab and switch port configuration changes should be submitted to [[servicedesk@redhat.com]]. Red Hat IT manages all switches in the Sepia lab and switch port configuration changes should be submitted to [[servicedesk@redhat.com]].
Line 54: Line 58:
 </​file>​ </​file>​
  
-==== Public IPs ====+===== Public IPs =====
 The Community Cage has a block of public IPs dedicated to it.  Ceph's usable IPs are 8.43.84.129 - 8.43.84.186. The Community Cage has a block of public IPs dedicated to it.  Ceph's usable IPs are 8.43.84.129 - 8.43.84.186.
  
Line 67: Line 71:
 | 8.43.84.136 ​ | [[production:​lists.ceph.io]] ​               | | 8.43.84.136 ​ | [[production:​lists.ceph.io]] ​               |
 | 8.43.84.137 ​ | [[production:​telemetry.ceph.com]] ​          | | 8.43.84.137 ​ | [[production:​telemetry.ceph.com]] ​          |
 +| 8.43.84.138 ​ | Telemetry public-facing read-only VM        |
 +| 8.43.84.139 ​ | [[production:​chacra.ceph.com]] ​             |
 +| 8.43.84.140 ​ | [[production:​beta.ceph.io]] ​                |
 +| 8.43.84.141 ​ | [[services:​quay.ceph.io]] ​                  |
  
-==== IPv6 ====+===== IPv6 =====
 IPv6 was enabled on 12/11/2018. IPv6 was enabled on 12/11/2018.
  
 Our subnet is ''​2620:​52:​3:​3:​0:​0:​0:​0/​64''​\\ Our subnet is ''​2620:​52:​3:​3:​0:​0:​0:​0/​64''​\\
 Our default gw is ''​2620:​52:​3:​3:​ffff:​ffff:​ffff:​fffe/​64''​ Our default gw is ''​2620:​52:​3:​3:​ffff:​ffff:​ffff:​fffe/​64''​
services/networking.1581612059.txt.gz ยท Last modified: 2020/02/13 16:40 by djgalloway