802.3ad on switches

Similar Questions

• Reference Dell PowerConnect 8024 - 802.3ad (dynamic LACP) through 2 switches?

  We evaluate the use of switches Dell PC8024 to upgrade our backbone to 10GbE.  We use currently HP ProLiant servers with multiple network cards are associated.  The servers support 802.3ad HP dynamic aggregation of links (LACP).

  The switch PC8024 does support the aggregation of dynamic links through 2 different switches, or should it be on the same switch?

  In comparison, the HP2910 switches can support aggregation (LACP) dynamic links across multiple switches, while the HP2920 switches can.  If some can, some can't.  Need to understand the ability of the switch to PC8024.

  Thank you!

  -Terry

  Hello Terry,

  The 8024 does not support a SHIFT of two different switches, the only way it would work, if the switches are stacked and effectively a single switch. If you have used the new series N4000 they take over MLAG, allowing for the OFFSET will be in both switches and they communicate via a peer.

• The GS324 supports 802.3ad Dynamics?

  I am looking to buy a Nighthawk R8500, and a switch of 24 ports of GS324 for my home network. However, I have found no information to indicate whether or not the GS324 supports double AD ports on the router R8500. It would really cinch the purchase for me, otherwise I'm on the fence.

  Does anyone know if the switch supports? If not, what switch would be recommended to get that extra bandwidth?

  Hello DasMainframe,

  Welcome to the community!

  The GS324 switch is only a (silent) switch passthrough and I do not think that it supports the LACP Protocol. It is not configurable, since she does not have any user interface or management software. If you want a switch with LACP (802.3ad) you have to go with our Smart Switch (PROSAFE series) models.

  Let us know if you have any other questions.

  Thank you

• 802.3ad - N4032 - 1679U-RP Qnap PowerConnect

  I need to get Qnap 1679U-RP with Qnap 10 G dual-port card (Intel X 520 based) to connect to stacked 2 x N4032 routing port 802.3ad

  On the Qnap there is not much available settings:

  Port Trunking & mode IEEE 802.3ad

  I have configured as a dynamic PO1 on the switch (s)

  #show console interfaces port-channel 1

  Ch-Type Hash Type Min-links Channel ports local FRP

  -------   ----------------------------- -------- --------- --------- ---------

  PO1 Active: 0/19/Te1, Te2/0/19 1 7 dynamic disabled

  Type of hashing algorithm
  1 - the source MAC, VLAN, EtherType, source module and port Id
  2 - destination MAC, VLAN, EtherType, module source and port Id
  3 - source IP and TCP/UDP source port
  4 - destination IP and destination TCP/UDP port
  5 source/Destination MAC, VLAN, EtherType, MODID/source port
  6 source/Destination IP and port TCP/UDP source / destination
  7 - improved hash mode

  But I have no connectivity at all

  SEB

  When you create the OFFSET on the mode of the switch channel group 1 on the Page 593 ftp.dell.com/.../networking-n4000-series_Reference%20Guide_en-us.pdf

  For server R2 2012 choose static grouping blogs.technet.com/.../nic-teaming-in-windows-server-2012-do-i-need-to-configure-my-switch.aspx

• 802.3ad link Aggregation - 4 NETWORK cards

  I'm working on a project that requires four physical NIC be connected to two separate Cisco switches for redundancy.  I need access to a minimum of six networks through these four physical NETWORK card ports using ESX 3.5.

  Our plan is to use the Routing and aggregation (Etherchannel) on Cisco switches.  What we want to do is basically aggregate two NIC physical port trunks to each of the two switches Cisco Etherchannel.  We would allow routing based on IP hash on the unique vSwitch with four network adapters are associated.  We would use a separate vSwitch for vSwif0 and vmkernel port not part of the trunks.

  Seems this configuration good to you?  Looking for networking people comment on the above configuration.  Thank you.

  This is a discussion that asked the same question.

  http://communities.VMware.com/docs/doc-5000

  This is an insertion of the blog of Kevin Cline.

  http://kensvirtualreality.WordPress.com/2009/04/05/the-great-vswitch-debate%E2%80%93part-3/

  A technical requirement for the use of hash of the IP, it's that your physical switch must support 802.3ad static aggregation of links. Often, this means that you have to connect all the natachasery in the vSwitch to the same pSwitch. Some switches high-end support links grouped in pSwitches, but many do not. Contact your switch vendor to find out. If you do not have to put an end to all the natachasery in a single pSwitch, you introduced a single point of failure in your architecture.

  If your provider supports the links grouped in all of pSwitches, great.  Otherwise, you may need to consider another approach.  You can have redundancy between switches without etherchannel and just routing based on originating port ID.  So you could create a vSwitch with 4 NICs (two on each pSwitch) without a trunk and let ESX base for outbound traffic load balancing.  From my experience, you are much more likely to be CPU or memory bound as bound network.

• DVS does support LACP 802.3ad?

  Hello

  Could someone let me know if DVS supports LACP (dynamic negotiation) 802.3ad. I know that vSwitch he has not supported.

  Thank you

  According to this document, only the Nexus 1000V supports LACP

  http://VMware.com/files/PDF/technology/cisco_vmware_virtualizing_the_datacenter.PDF

  There is a table that shows which features are available on each type of switch.

• Cisco 1252 s with regular 802.3af PoE switch - what are the impacts?

  Does anyone have an experience of real life to the deployment of the 1252 s on 802.3af regular PoE switches?  I have a few 1252 s a certain ports PoE 3560 and issuing upward with alarms 'low power '.

  I wonder what the impact of the real world with this scenario.  The 1252 s have both 2.4 Ghz and a radio-5, 0 Ghz in them and I have HT enabled on the controller.

  I wonder what capabilities I lose by running on 802.3af food in addition to the inconvenience of the alarms.

  Thanks in advance!

  Hello

  This is the command question... the AP 1252 switches POE needs improved Ex-3750-E for all operational radios... If we use only a 802.3af while a single frequency can be enabed and her we can't get a rhythm N... the link below will explain better...

  http://www.Cisco.com/en/us/prod/collateral/wireless/ps5678/ps6973/ps8382/product_data_sheet0900aecd806b7c5c.html

  I would like to know how it works for you!

  Concerning

  Surendra

• 802.1 x Switch IOS Bug?

  Seems to me IOS does not work as documentation States during the dot1x authentication management. In my view, that the last IOS 12.2 should not authenticate a client if the MAC address has not changed (with dot1x reauth off of course). However, I've tested this and it seems that the switch always sends EAPOL, even if I use the same PC on the same port. Is this a bug?

  There is nothing that can be done about it. I don't know yet what means 'History of MAC', but if the port goes down, it is still deleted.

  Now, I might have a work around for you here. MAC-Auth-Bypass (MAB). MAB authenticates the machines that cannot speak 1 X be their MAC address. If it fails and you have also the Guest-VLAN on, the port is being Guest-VLAN anyway (to support backward compatibility). From a treatment perspective, MAB is attempted after 1 X, but before the comments - VLAN (who just allow a port blindly).

  What it means for your scenario here is that if you enable MAB, you can put a machine to sleep that will bounce the port). 802. 1 X expires (as your computer is idle). Then, MAB will kick and initiate. However, it will be hung there until the device is sending traffic, and while it is asleep, it will not pass the. In this way, the port does not in the comments - VLAN when go to sleep and you can wake the computer that upward of a VLAN is configured in native mode on the port.

  Hope this helps,

• Safe on Etherchanel / 802.3ad, it works with vmware esx?

  Trunk etherchanel in cisco switch configuration, it works with vmware esx?

  Configuration of the Cisco switch like that (I don't have a switch, can not test it):

  src-dst-ip port-channel load-balance
  the interface vlan 4094
  no ip address
  Interface Port - Channel 1
  switchport
  switchport mode trunk
  switchport trunk encapsulation dot1q
  switchport trunk allowed all the VLANs
  switchport trunk vlan 4094 native
  no ip address
  interface GigabitEthernet1/1
  switchport
  switchport mode trunk
  switchport trunk encapsulation dot1q
  switchport trunk allowed all the VLANs
  switchport trunk vlan 4094 native
  no ip address
  Chanel-Group 1 mode on
  interface GigabitEthernet1/2
  switchport
  switchport mode trunk
  switchport trunk encapsulation dot1q
  switchport trunk allowed all the VLANs
  switchport trunk vlan 4094 native
  no ip address
  Chanel-Group 1 mode on
  end

  What you have should work OK, because it is not too different from the example of work from my lab:

  ocs4948-1 #sh int item in gi1/2nd round

  Building configuration...

  Current configuration: 536 bytes
  !
  interface GigabitEthernet1/2
  switchport trunk encapsulation dot1q
  switchport trunk allowed vlan 2.10
  switchport mode trunk
  MTU 9198
  channel-group 64 mode on
  spanning tree portfast
  spanning tree enable bpduguard
  end

  ocs4948-1 #sh run int item in gi1/6

  Building configuration...

  Current configuration: 536 bytes
  !
  interface GigabitEthernet1/6
  switchport trunk encapsulation dot1q
  switchport trunk allowed vlan 2.10
  switchport mode trunk
  MTU 9198
  channel-group 64 mode on
  spanning tree portfast
  spanning tree enable bpduguard
  end

  ocs4948-1 #sh run int in64

  Building configuration...
  Current configuration: 291 bytes
  !
  Interface Port-channel64
  Description esx402 LACP link
  switchport
  switchport trunk encapsulation dot1q
  switchport trunk allowed vlan 2.10
  switchport mode trunk
  MTU 9198
  load-interval 30
  Storm-control broadcasts 5.00
  spanning tree portfast
  spanning tree enable bpduguard
  end

  ocs4948-1 #sh int trunk

  VLAN Mode Encapsulation native port State
  In64 on 802. 1 q trunking 1

  Port VLAN allowed on trunk
  In64 2.10

  Port VLAN authorized and active in the field of management
  In64 2.10

  VLAN port extending on transmission State and no tree pruned
  In64 2.10

  The only thing to look at will be the VLANs that you got set up on the vSwitch. You have defined VLAN native example, the VLAN untagged on the switch to be 4094, any traffic sent on the aura tags trunk VLAN, which will be removed by the vSwitch.

  In my setup, I have established groups of ports using VLANS and VLAN 2 10. Traffic to the virtual machines on these VLANs are .1Q marked by the physical switch, which is then stripped by the vSwitch.

  Take a look at the section "VLAN in VMware Infrastructure" on page 7 the concepts of virtual networking André already mentioned, but also the other interesting reading material is http://www.vmware.com/pdf/esx3_vlan_wp.pdf.

  Finally, for the side of the switch of things, look at the examples on Scott Lowe's blog at http://blog.scottlowe.org/2006/12/04/esx-server-nic-teaming-and-vlan-trunking/.

  Concerning

• Packed SLM2008 fall

  Hello

  I recently looket in the following thread: http://homecommunity.cisco.com/t5/Switches/Linksys-SRW2008-Load-balance-setup/td-p/217487

  I recently posted this on the QNAP forum: http://forum.qnap.com/viewtopic.php?f=50&t=56416

  I don't know where the problem is, but maybe someone here could give it a glance and a few tips. Protocol I run the 802.3ad and switch SLM2008 LACP configuration. You can see the pictures here: http://www.dropmocks.com/mBg-X8

  The MTU is set to 9000 on the servers and the SLM2008 Jumbo Frame active.

  Sender-Server(ts-859)
  
  Reciever-Server(ts-639)
  
  Switch configuration
  

  Hello.

  These products are processed by the Cisco Small Business Support Community.

  * If my post answered your question, please mark it as "acceptable Solution".

  Thank you!

• LACP y switch virtual Standard

  Hola

  Sabeis if is pueden configurar a virtual standard switch con back tarjetas para as funcionen in el switch fisico configurado con LACP aggregation?

  decir, if tenian el grouping?

  Gracias.

  Hola compañero,

  a Día de hoy solo supports vmware's para el uso en version of vSphere 5.1 LACP solo VDS (VMware vDistributed rocking):

  http://KB.VMware.com/selfservice/microsites/search.do?cmd=displayKC&docType=kc&docTypeID=DT_KB_1_1&externalID=2034277

  In the rest of scenarios, para hacer NIC teaming is debe usar (para cisco switch) Etherchannel con el LACP (802.3ad) desactivado. Modo in "static mode WE:

  http://communities.VMware.com/message/902563#902563

  Te dejo par UN of vmware as explican muy KB well el tema of NIC teaming en vSphere:

  http://KB.VMware.com/selfservice/microsites/search.do?cmd=displayKC&docType=kc&externalId=1001938&sliceId=1&docTypeID=DT_KB_1_1&dialogID=675078772&StateID=1%200%20675098848

  http://KB.VMware.com/selfservice/microsites/search.do?cmd=displayKC&docType=kc&externalId=1004048&sliceId=1&docTypeID=DT_KB_1_1&dialogID=675078772&StateID=1%200%20675098848

  Saludos,

  Pablo

• Best configuration of switch for 4 natachasery - 5 ESXi

  So, I need to tell my boys to network how I want the configuration of ports (Cisco switches) for 4 natachasery connection to my HP DL 360. I need the vSwitch VM access to 3 different VLAN. I'm NOT going to use vMotion for this.

  I guess, I see 2 possibilities and does not entirely undesrstand pros / cons...

  First of all, I thought I would split the enet to 2 separate switches cables, but I was told that if I chose that this channel of ports 1 route would be active and 2nd would be Eve. Basically can not paste all 4 cards and divide them. This could reduce the bandwidth, but give me a redundancy using 2 physical switches.

  Or I can have all 4 enet cables go to the same switch (etherchannel), but if this switch would die I'm dead in the water, AND I think that EtherChannel mode WE (my network guy preferred LACP) should stay and LACP cannot be used. According... http://KB.VMware.com/selfservice/microsites/search.do?cmd=displayKC & docType = kc & docTypeID = DT_KB_1_1 & externalId = 1004048 I think I should also use 'Hash IP' if EtherChannel is used.

  So what is the advantage or disadvantages of each channel? Thank you!

  This article describes the aggregation of ESX/ESXi links:

  • ESX/ESXi host only supports grouping NETWORK adapters on a single physical switch or stacked switches.
  • Aggregation of links is never supported on disparate networks switches.
  • The switch must be set to perform 802.3ad aggregation of links in static mode WE and the virtual switch must have its defined route based on the IP hashload balancing method.  Ensure that participants of the network adapters are connected to the ports configured on the same physical switch.
  • Allowing each route based on the IP hash without 802.3ad aggregation or vice versa disrupts networking, so you must make the changes to the virtual switch first. In this way, the service console is not available, but the physical switch management interface is, so you can activate the aggregation on the affected ports to restore network.
  • For more information on ESX networking concepts, see the VMware Virtual Networking Concepts.
  • For more information on configuring a VLAN for NIC teaming, see Configuration of VLAN Virtual Switch, switch physical and virtual machines (1003806).
  • To set up the consolidation of NETWORK cards on ESX and a physical switch, see example configuration of EtherChannel / link aggregation with ESX/ESXi and Cisco/HP switches (1004048).
  • The only option for switch vSwitch or vDistributed that can be used with balancing EtherChannel is HASH IP.
  • Do not use tag with IP HASH load balancing.
  • Do not uplinks configure standby or not well obliterated with IP HASH load balancing.
  • VMware takes support only a single Etherchannel by vSwitch or distributed vNetwork (vDS).

  heybuzzzz wrote:

  In my case with no vMotion and just management and 3 VLAN all about 1 vSwitch I should put them all to active?

  Yes, they should all be active, there is really no need passive adapters (that I'm aware of your configuration).

  On your physical switches, you don't need to do anything more than what allows a VLAN correct, i.e.: should all etherchannel or a similar montage. Side vSwitch, you must use the default 'Port ID' nic teaming strategy.

• External Tagginng switch vs Virtual Switch Tagging

  Hello guys,.

  I'm really confused about these modes of two configurations and their uses / consequences.

  Say you have four natachasery and configure the vSwitches in the following way:

  • a vswitch with service console port group (vmotion) vmkernel port group and group of ports defined VM and four natachasery in annex

  All three services (SC, vmotion and vmtraffic) are in the same network (e.g.: 192.168.1.X), so all the natachasery is attached to the physical integrity go to this VLAN.

  In this scenario, I guess I'll use IS, right? I have to configure something in the physical switch (like the 4 natachasery trunking)? Or just nic teaming four cards in the vswitch network is enough? Or both?

  =======================================================================================================

  Now say you have 6 natachasery and configure the vSwitches in the following way:

  • a vswitch with defined service console port group and two natachasery in the annex

  • a vswitch with Group of ports vmkernel (vmotion) and two natachasery defined in annex

  • a vswitch with Group of vm ports and two natachasery defined in annex

  Each service has its own network, but now you have the production and DMZ VM.

  In this scenario, I guess I'll have to do the following (get ready cause its really confusing, at least for me):

  (1) set two NICs in the pswitch, firstly in the vlan management and in the nic vswitch team these two network cards. If you use felt

  (2) attach the second pair of NICs in the pswtich, in the vlan vmotion and nic vswitch team these two network cards. If you use felt

  (3) fix the last pair of network adapters in the pswitch to two ports that don't have any definied vlan on it and then to create two groups of ports in the vswitch: one for the VM of the production and the other for VM DMZ, marking with the right VLAN ID and nic team these two network cards. If you use the VST.

  Edit: I realized that I could simplify the second scenario, saying that the three services would be in the same vswitch with 6 natachasery attached to it, but with four networks. And fix these 6 cards in the pswitch with no VLAN defined in these 6 ports.

  As you can see, I'm pretty new in this kind of things, so, if I'm wrong to speak of something (or all), I'm sorry.

  I have attached two Dummies to try to be more clear.

  Excuse me for my English

  Post edited by: brian_plank

  In the first scenario, you are right.  No obligation to trunk.  On the other hand, I myself prefer to use trunks for all my connections, with all of my virtual local networks available on all of my trunks.  In this way, if I need to spend my autour natachasery, ever, I can do without the participation of network.  If you yourself control the network, then this isn't a problem.  Of course, this is not required, the way you have your first scenario presented will work fine.  And by the trunk, I mean a 802. 1 q trunk in terms of cisco, as opposed to a trunk in terms of HP, which would be an aggregation of links 802.3ad.

  In scenario 2, only the vSwitch vm network exchanges where you need multiple VLANs is where you need the trunk.  Make sure just that the VLAN native on this trunk is different.  In addition, you need to configure the ports as an access port (which means that the port will be part of a single VLAN) for the management VLANs, or a trunk with several port VLAN allowed.  Other than that, there should be no problem with VLAN marked and unmarked.  I use them all the time.

  -KjB

  VMware vExpert

• AirPort Extreme, gigabit Ethernet switch and covering my house

  I have an AirPort Extreme 802.11n to Time Capsule wireless router. I have an old House. It was built in 1852 and microwave ovens have a hard time penetrating through some walls.

  Recently, I made some changes in the House and I wired things for Gigabit Ethernet. I have not yet bought a switch. We have a Mac Pro (cheese grater), a Mac Mini, three iPhones (only 6) and two iPads (one version 3 and the other, only the latter bought autumn, an iPad Air.)

  First of all, since most of us are on 802.11 n, can I initially put in place a heterogeneous system with two routers wireless connected to themselves and to the Internet via a switch? Also, I need two switches; one for the internal LAN and the other for a connection to the Internet? I intend to buy the new ac 802.11 AirPort extreme for a facility in one of the rooms to better cover the rest of the House. I think that both devices will be able to take advantage of the signal of 5 GHz, initially, with another iPad and iPhone to come along the next year. I wouldn't or the other base station being "enslaved" to the other but, rather, connected to the wired network internal and devices just switch between them that move around the House.

  Comments are welcome here. This can be a trip to a Genius Bar if I do not hear from final here.

  can I initially put in place a heterogeneous system with two routers wireless connected to themselves and to the Internet via a switch?

  Possibly, depending on the type of modem you have. See below info in this post.

  Also, I need two switches; one for the internal LAN and the other for a connection to the Internet?

  NO.

  I wouldn't or the other base station being "enslaved" to the other but, rather, connected to the wired network internal and devices just switch between them that move around the House.

  Not sure what you mean here. The two extremes of the airport will be connected to the cable network, so they will be in communication with them at all times. If you want to peripheral switch between them that move around the House... then you will need what is called a "mobile network"... so the two airports will produce a wireless signal that uses the same wireless network name. Apple Setup Assistant will put things this way by default.

  Depending on the type of "modem" you have, the network might look like this:

  Internet > Modem > AirPort Extreme 802.11ac > Ethernet switch > AirPort Extreme

  or, if your "modem" is really a modem/router, also known as, a walkway... then you have the ability to set things up like this:

  Internet > Modem/Router > Switch > the two airports to connect to the Ethernet network

  Looking for a Gigabit Ethernet switch to your network. Switches come in 5 ports, 16 channels, 8 ports, etc. versions.  It's always a good idea to buy an Ethernet switch over a few ports that you think that you may need right now... to allow for expansion.

  The key here is that the type of modem you have will determine how to configure the network.  If you could provide the number of brand and model of your "modem", we can eliminate some of the assumptions to be used at this time.

• 802. 1 x with dACL - prefix of an invalid attribute: "ACS."

  Dear all,

  I spent half an update to fix this problem without success, I hope you could help me.

  I configured a simple solution of 802. 1 x on a PC driver who must authenticate through PEAP-MSCHAPv2 users against my user database internal GBA.

  Version of the switch:

  Model number: WS-C3750V2-48PS-S

  Software: c3750-ipbasek9 - mz.122 - 52.SE.bin

  ACS:

  C1121 with version 5.3.0.40

  The problem occurs when the ACS sends within the radius Authentication accept packet the following attribute:

  Cisco-AV-pair=ACS:CiscoSecure-defined-ACL=#ACSACL#-IP-auth-4eb90704

  On the side of the switch, I see the following debug log:

  002558: 8 Nov 14:31:35.586: % AUTHMGR-5-START: start "dot1x' for the client (0022.680b.da7b) on the Interface Fa1/0/1 AuditSessionID AC1FFE4E0000003105BCDE19

  002559: 14:31:35.703 8 Nov: AAA/ATTR: prefix of an invalid attribute: "ACS."

  002560: 8 Nov 14:31:35.703: % DOT1X-5-FAIL: failure of authentication for the client (0022.680b.da7b) on the Interface Fa1/0/1 AuditSessionID AC1FFE4E0000003105BCDE19

  002561: 8 Nov 14:31:35.703: % AUTHMGR-7-RESULT: result of the "dead server" authentication of 'dot1x' for the client (0022.680b.da7b) on the Interface Fa1/0/1 AuditSessionID AC1FFE4E0000003105BCDE19

  802.1 x switch associated config:

  GLOBAL:

  Group AAA dot1x default authentication RADIUS

  Group AAA authorization network default RADIUS

  start-stop radius group AAA accounting dot1x default

  RADIUS-server host 172.31.254.140 auth-port 1645 acct-port 1646

  RADIUS-server host 172.31.254.141 auth-port 1645 acct-port 1646

  RADIUS server key 7 123415ASFASFAS55512

  RADIUS vsa server send accounting

  RADIUS vsa server send authentication

  analysis of IP device

  IP access-list extended by DEFAULT, ALL

  allow an ip

  SPECIFIC PORT

  interface FastEthernet1/0/1

  Description model Port 802. 1 x

  switchport access vlan 244

  switchport mode access

  IP access-group by DEFAULT, while

  authentication event fail following action method

  open authentication

  authentication priority dot1x mab

  Auto control of the port of authentication

  periodic authentication

  MAB

  dot1x EAP authenticator

  dot1x tx-time 10

  end

  Next to the ACS authentication ends successfully, but for some reason, the switch cannot understand attribute was sent by the ACS:

  Why Authentication translates as 'server-dead?

  Hereby, I have attached the authorization profile, the downloadable ACLs and the detail of the RADIUS authentication for the request...

  Any idea?

  Thank you very much!

  Yes, I came across the same issue and ended up as a bug with the 3750

  CSCtj28883 dACL attribute the parsing failed when debug "author of aaa" on

  Description is

  The DACL processing fails when the following debug settings are turned on.

  1 debug aaa attr

  2 debug aaa authorization

  The same works very well when they are turned down. Set the switch of newspaper.

  I believe has been resolved in version 3750-Build 12.2 (55) as to the next note, attached to the bug as proved to be irreparable on later constructions

  The issuer has confirmed that the bug is not seen on the image of 55SE.

  The issue is only seen in 53SE

  can also try and switch debug off