Network design

From Wikitech
(Difference between revisions)
Jump to: navigation, search
(Florida cluster (PMTPA): - snap the link to PM)
(+links)
 
(23 intermediate revisions by 8 users not shown)
Line 1: Line 1:
The purpose of this page is to give an overview of the current '''design of the network''' of the Wikimedia servers, and to provide a place to develop a new and improved network scheme.
+
== AS 14907 ==
 +
The US network.
  
 +
=== 2011 ===
 +
[[File:Eqiad logical.png|thumb|400px|AS14907 Eqiad in 2011]]
 +
[[File:Wikimedia pmtpa management network.png|thumb|400px|AS14907 in 2010]]
  
==Overall system design==
+
=== Subnets ===
The folowing is the general system design plan which the network layer must efficiently accommodate.
+
  
*Databases in a central pool with each serving a subset of the wikis, so each has high cache efficiency and the total number needed to handle any query load is minimised. Database servers cost US$5,000-$8,000 each, depending on exact equipment.
+
==== [[pmtpa]] ====
*A central pair of old text database servers (part of the long term storage growth plan for the databases, to move this high volume and seldom accessed data off costly and comparatively small disk systems).
+
*Memcached caching spread on apaches across the whole cluster, producing one very large cache pool, accessible from any apache and stored on half or more of the apaches. Segmenting the pool would decrease the overall hit rate, increasing the number of apaches and database servers required for any given system load level.
+
*Load balancing of squids and apaches, currently expected to use two or three systems between the internet and the squids and the same set between the squids and the apaches.
+
  
A key network systems design requirement is efficient access from any apache to any apache running memcached (expected to be more than half of all apaches) and efficient access from any apache to any database server. Losing this capability would dramatically increase overall system cost.
+
==== [[eqiad]] ====
 +
{| class="wikitable"
 +
|-
 +
!subnet          !! vlan ID !! IPv4 !! IPv6
 +
|-
 +
| public1-a-eqiad  || 1001 || 208.80.154.0/26  || 2620:0:861:1::/64
 +
|-
 +
| public1-b-eqiad  || 1002 || 208.80.154.128/26 || 2620:0:861:2::/64
 +
|-
 +
| public1-c-eqiad  || 1003 ||                  || 2620:0:861:3::/64
 +
|-
 +
| public1-d-eqiad  || 1004 ||                  || 2620:0:861:4::/64
 +
|-
 +
| private1-a-eqiad || 1017 || 10.64.0.0/22      || 2620:0:861:101::/64
 +
|-
 +
| private1-b-eqiad || 1018 || 10.64.16.0/22    || 2620:0:861:102::/64
 +
|-
 +
| private1-c-eqiad || 1019 || 10.64.32.0/22    || 2620:0:861:103::/64
 +
|-
 +
| private1-d-eqiad || 1020 || 10.64.48.0/22    || 2620:0:861:104::/64
 +
|}
  
== Current situation ==
 
  
The cluster names are a two character code for the colo provider, and a 3 character code for the city location.  (Candidhosting is now Power Medium :-).
+
== AS 43821 ==
 +
The European network.
  
=== Florida cluster (PMTPA) ===
+
=== 2008 ===
Wikimedia servers reside in two racks along with Bomis servers, hosted at [http://powermedium.com/ PowerMedium]. Wikimedia/Bomis have a dedicated IP range, <tt>207.142.131.192/26</tt>. There are two gateways: <tt>207.142.131.193</tt> and <tt>207.142.131.225</tt>. Total burstable bandwidth is 1000 Mbit/s, delivered through one optic fiber 1000base-SX link.
+
[[File:Knams-multihomed.png|thumb|400px|AS43821 in 2008]]
  
Wikimedia owns a Cisco C3560G-48TS switch, and three Netgear Gigabit [[switches]].
+
BGP default transit from AS1145 (Kennisnet), with some partial transit and peering over a 1 Gbps AMS-IX link. Everything on one core router/switch, csw1-knams (Foundry BigIron RX-8).
  
=== Amsterdam cluster (KNAMS) ===
+
=== 2009 ===
[[Image:Knams-network.png|thumb]]
+
[[File:AS43821 2009.png|thumb|400px|AS43821 in 2009]]
The Kennisnet cluster's network follows a similar design as the Florida cluster, with one Cisco 3560G-24 core switch connected via a Gigabit ethernet port in ''routed'' mode to the uplink Kennisnet router. The L3 switch routes between "The Internet" and the Wikimedia VLANs:
+
; VLAN 100 : Public VLAN (<tt>145.97.39.128/27</tt>)
+
; VLAN 101 : Private VLAN (<tt>10.20.1.0/24</tt>)
+
; VLAN 10 : Installation VLAN
+
  
There is a separate ''management network'' (<tt>145.97.34.224/29</tt>), with a separate uplink connected to a Kennisnet firewall for [[Kennisnet out of band access|out of band access]]. It's connected to the Service Processors of all Sun servers in a daisy chain.
+
Temporary situation after the move from knams to esams. The network is split, with a new Foundry BigIron RX-4 as a pure router at knams for external connectivity, with Telia, DataHop, Init7 (partial) transit, and 2x 1 Gbps AMS-IX for peering. Connectivity between the two sites is supplied by a 10GBase-ER link over dark fiber, and a 3 Gbps MPLS backup link. A second dark fiber is being installed to form a ring.
  
[[Category:Network]] [[Category:Kennisnet cluster]]
+
 
 +
=== 2010 ===
 +
[[File:AS43821 Q3 2010.png|thumb|400px|AS43821 late 2010]]
 +
 
 +
The purchase of several Juniper EX4200s in a stack, for extra access ports for servers, also brings some opportunities w.r.t. the network topology. Since the EX4200s have excellent L3 support they can help create redundancy.
 +
 
 +
The 2nd dark fiber is linked between [[br1-knams]] and [[csw2-esams]] to create a ring. [[csw1-esams]] and [[csw2-esams]] can then share responsibility as core switches, for inter-vlan routing and switching, using VRRP. Since an EX4200 can not install a full BGP routing table in FIB, it defaults to either of the two Foundry routers using OSPF.
 +
 
 +
Toolserver can be connected redundantly as well, using (R)STP to both core switches and VRRP, or alternatively a LAG to the EX4200 stack.
 +
 
 +
== Configuration guidelines ==
 +
* Firewall filters, policies, prefix lists etc that are specific to a certain protocol family (e.g. only IPv4, or only IPv6) should have a '4' or '6' appended to their name. Filters, policies and prefix lists that are protocol family agnostic, should lack this suffix.
 +
 
 +
== See also ==
 +
* [[Multicast]]
 +
* [[TCP Tuning]]
 +
 
 +
[[Category:Network]]
 +
[[Category:knams cluster| *]]
 +
[[Category:Pmtpa cluster| *]]

Latest revision as of 22:43, 20 February 2012

Contents

[edit] AS 14907

The US network.

[edit] 2011

AS14907 Eqiad in 2011
AS14907 in 2010

[edit] Subnets

[edit] pmtpa

[edit] eqiad

subnet vlan ID IPv4 IPv6
public1-a-eqiad 1001 208.80.154.0/26 2620:0:861:1::/64
public1-b-eqiad 1002 208.80.154.128/26 2620:0:861:2::/64
public1-c-eqiad 1003 2620:0:861:3::/64
public1-d-eqiad 1004 2620:0:861:4::/64
private1-a-eqiad 1017 10.64.0.0/22 2620:0:861:101::/64
private1-b-eqiad 1018 10.64.16.0/22 2620:0:861:102::/64
private1-c-eqiad 1019 10.64.32.0/22 2620:0:861:103::/64
private1-d-eqiad 1020 10.64.48.0/22 2620:0:861:104::/64


[edit] AS 43821

The European network.

[edit] 2008

AS43821 in 2008

BGP default transit from AS1145 (Kennisnet), with some partial transit and peering over a 1 Gbps AMS-IX link. Everything on one core router/switch, csw1-knams (Foundry BigIron RX-8).

[edit] 2009

AS43821 in 2009

Temporary situation after the move from knams to esams. The network is split, with a new Foundry BigIron RX-4 as a pure router at knams for external connectivity, with Telia, DataHop, Init7 (partial) transit, and 2x 1 Gbps AMS-IX for peering. Connectivity between the two sites is supplied by a 10GBase-ER link over dark fiber, and a 3 Gbps MPLS backup link. A second dark fiber is being installed to form a ring.


[edit] 2010

AS43821 late 2010

The purchase of several Juniper EX4200s in a stack, for extra access ports for servers, also brings some opportunities w.r.t. the network topology. Since the EX4200s have excellent L3 support they can help create redundancy.

The 2nd dark fiber is linked between br1-knams and csw2-esams to create a ring. csw1-esams and csw2-esams can then share responsibility as core switches, for inter-vlan routing and switching, using VRRP. Since an EX4200 can not install a full BGP routing table in FIB, it defaults to either of the two Foundry routers using OSPF.

Toolserver can be connected redundantly as well, using (R)STP to both core switches and VRRP, or alternatively a LAG to the EX4200 stack.

[edit] Configuration guidelines

  • Firewall filters, policies, prefix lists etc that are specific to a certain protocol family (e.g. only IPv4, or only IPv6) should have a '4' or '6' appended to their name. Filters, policies and prefix lists that are protocol family agnostic, should lack this suffix.

[edit] See also

Personal tools
Namespaces

Variants
Actions
Navigation
Ops documentation
Wiki
Toolbox