Interconnection of fabric as FC switch

Similar Questions

• Connection of fabric interconnect with the Cisco MDS switch

  I'm new to the unified fabric and try to understand it.

  There are clients who connect fabric connects directly to switch MDS without any Nexus 5 K between the two.  If the fabric of interconnections are able to provide connectivity to the LAN and SAN Islands then why you install Nexus 5 K between two switches to add an extra jump.

  Thank you

  Hello

  Most of the clients do the following:

  -connect FI upstream for example N5k with vPC for Ethernet (IP)

  -connect FI upstream for example MDS for the classic FC

  This implies that between the blade into the chassis and FI you run FCoE, integrating CF and Ethernet on the same means of transport; then, on the FI, FC and Ethernet is disintegrated.

  The other choice would be Multihop FCoE, linking FCoE also FI and N5k and make the disagregation it.

  However, if your storage subsystem is FC and FCoE, I wouldn't recommend Multihop FCoE (I see only a minority of clients to do).

  Walter.

• Adding ports to a vlan on interconnection of fabric (end-host Mode)

  Hi all

  I have a problem, I have created vlan 99 and added the ports and port channels to the vlan on the fabric of the ucs interconnections and saved the configs.

  To my surprise, that nothing has changed, they are still on the vlan 1 (the default).

  See my setup below.

  UCS - A eth-link # rising scope
  UCS - Has/eth-uplink # create vlan EMM_NET 99
  UCS - Has/eth-storage / vlan * # create member-port has 15 1
  UCS - Has/eth-storage / vlan * # create member-port a 16 1
  UCS - Has/eth-storage / vlan * # create portchannel members a 61
  UCS - Has/eth-uplink / vlan * # set no sharing
  UCS - Has/eth-uplink / vlan * # commit buffer
  UCS - Has / eth-uplink/vlan #.

  Or should I just create vlan 99 on the switch to the North (Nexus5548UP) because the fabric of interconnection is in fine host mode?

  Kind regards

  OK, now you must give us more detailed information?

  What is your OS? ESXi?

  And maybe, depending on how your host is configured, you should make 99 Vlan as vlan native!

• UNI-DIRECTIONAL press UCS 6248 interconnections of fabric?

  Hello

  If possible, I'd like to see a UNI-DIRECTIONAL between my 5548 neighbor was cheating Nexus and my 6248 interconnection fabric.

  UNI-DIRECTIONAL feature is supported by the 5548 s and easily activated.

  However, I can't find any information on how to activate on the UCS FI ports uplink ethernet. Is this possible?

  Thank you.

  Hello

  UNI-DIRECTIONAL UCS FINANCIAL support will be available in the next version major i.e UCSM 2.2

  UNI-DIRECTIONAL feature is currently not available on the FI.

  HTH

  Padma

• Name of the system of the interconnections of fabric can be changed

  I need to change the name of the fabric of interconnection system

  I gave for example. EDCD

  the IFS have now EDCD-A

  EDCD-B

  I need to change it to... PEPE-A

  PEPE-B...

  is this possible?

  Shyam,

  You can certainly do it and for sure are not disruptive and take effect immediately.    Simply go to the Admin tab > all > Management Interfaces > name > edit with you > Apply.

  

  It is as simple and quick as if at the same time, you have a SSH session, as soon as you change the name, the system made the change on your SSH session also.

  Note: There is no need to look for the way to do this on every fabric. The name you are editing is for both financial institutions and the system then adds the XYA -has or XYA -B by itself.

  Note THE answers useful.

  -Kenny

• The question about the interconnections of fabric UCS 6300 Series

  Hello

  I have a question about UCS 6300 Series fabric interconnects. Is this fabric UCS interconnections 6300 series integrated into the extension of tissue function?  The reason why I asked this question was, when there is 6300 series fabric has been plugged UCS 5100 chassis, there is no more space to plug in extension of UCS 2200 series fabric. My understanding on the topology of the UCS, tissue extender is the connection between the blade and the fabric of the interconnections.

  Thank you

  Bob

  Hi Bob

  Yes, that's correct!

  The chassis so called FI - IOM (see attachment p52); THE IOM = FEX.

  I/O of the blade adapter to connect inside basket chassis, which connects to the built-in IOM.

  IOM is then connected to the built-in FI.

• Why the interconnections of fabric can be connected to the two MDS?

  Hi all

  1. why the white paper below says that if FI is connected to two MDS, then it will create a single point of failure?

  http://www.Cisco.com/c/en/us/products/collateral/storage-networking/MDS-9500-series-multilayer-directors/white_paper_c11_586100.html

  This image below is used to describe the scenario for question 2 to 4.

  San.PNG

  

  2 and why the SAN design best practices to create SAN A and SAN B? Where when the switch 1 has failed, then it will use the b in SAN. And in this case, user can configure to connect to the storage using SAN B first, instead of using SAN first? And by default, the behavior of the SAN like that, what I mean is that she will check SAN failed, then it will use SAN B?

  3. and why there is no connection between the left more server and switch 2? They can not who design when switch 1 has failed, then it will use the other link first to switch 2, by linking the left more server for 2? Then when the core SAN switch one down, he uses the SAN B connection.

  4. can you the server to different SAN multipath? In this case, SAN A and B.

  Thank you

  Osbin

  Hi Osbin,

  Why the white paper below says that if FI is connected to two MDS, then it will create a single point of failure?

  Protocol FC build tissue, using the FSPF (such as OSPF) of the Protocol in order to MDS, FI (switch mode) Nexus 5 k, 7 k (storage VDC) form a single fabric using the links intercommutation.

  If a problem is occur in tissue alone device's problem (zoning, FSPF, EENT)

  If you connect fabric and fabric B (connect a FI for switch 1 and switch 2) in other words, this is no fabric A and B - it became a unique fabric (fabric single = single point of failure). That's why we use two FI, connected to two independent fabrics (isolation and redundancy)

  Technically, you can avoid this problem by using VSANS, but if some problem in a NX - OS or SAN - OS occure you may lose the two fabrics in a single failure (the worst of the world SAN)

  2 and why the SAN design best practices to create SAN A and SAN B? Where when the switch 1 has failed, then it will use the b in SAN. And in this case, user can configure to connect to the storage using SAN B first, instead of using SAN first? And by default, the behavior of the SAN like that, what I mean is that she will check SAN failed, then it will use SAN B?

  Point main initiator is in the answer above, can use both at the same time (multipath) fabrics with special drivers.

  It is configurable by the user to use any fabric (A and B - is just for engineers) first.

  It depends on the OS and drivers.

  3. and why there is no connection between the left more server and switch 2?

  I believe that this picture is schematic diagram and imagine if you have enough ports in the switch 1

  They can not who design when switch 1 has failed, then it will use the other link first to switch 2, by linking the left more server for 2?

  Remember the idea of FC fabric. If you spend 1 it will affect every single switch fabric, and it is much faster to use stable fabric B instead of A fabric converging.

  Then when the core SAN switch one down, he uses the SAN B connection.

  4. can you the server to different SAN multipath? In this case, SAN A and B.

  Yes

  Best regards

  Alex

• Passthrough interfaces and fabrics M1000e - M6220 switch?

  Hi all

  I have a M1000e with 4 switches x M6220 available, and I have also 2 x passthrough available modules.  I have 3 x FH M710 blades with fab A, B and C.  I'm also 5 blades FH x M905 with only A Fab and the B.  He would be allowed to use the Passthrough Fab B1 and B2 and switches in Fab A and C modules?    I realize that I won't be able to access Fab C blade M905 but they are for test and dev, so this isn't a problem.

  So is it ok to mix or is it not a good thing?

  Thank you

  Eric

  Hi all

  I have advanced and mixed modules and it works a treat for me.

• Fabric of interconnection Port Mode

  During our first installation of our UCS deployment only two of our rising four that connect the fabric to the frame were connected. This was done by a contractor. Now, we have two other cables and get them connected but mode on ports are set to "Trunk" and "tissue" as not the rest. The role has been defined in the interface of the UCS, but fashion was not so easy. How can I change this way of fabric?

  Hi Dennis,

  But there is NO need to be set to 4. As mentioned, the chassis discovery policy determines what is the minimum number of links for a frame be discovered/usable. See you to have it set to 2, it is very good and you can keep it this way.

  The fact that you have only 3 cables from IOM to the interconnection of fabric, this is not a valid configuration, because only 1, 2 or 4 links are taken care of. You must increase the number of cable to 4 and then re - emphasize the chassis to activate the additional links.

  Let me know.

  Thank you

  Michael

• Connectivity between Interconnect and Nexus 2000 fabric options

  Hello

  I would like to know what models of fabric nexus 2000 Extenders can I use if I want to connect them to a pair of interconnections 6248UP fabric in a FEX deployment, so I can move between blade and rack servers service profiles.

  Thanks for your replies.

  Christian

  Christian,

  Take a look at this guide.  It should answer all your questions:

  http://www.Cisco.com/en/us/customer/docs/unified_computing/UCS/c-series_integration/ucsm2.0.2/b_UCSM_202_C-integration.html

  Kind regards

  Robert

• FAS3220 and FAS2240 directly set on fabric of interconnection 6296UP

  Hello everyone

  I have the following architecture

  1 cisco Catalyst 6509-E

  1 cisco Catalyst 3560-x

  2 fabric of interconnection 6296UP

  1 FAS 3220 (primary storage)

  1 FAS2240 (backup of the vault)

  1 chassis cisco UCS 5108

  5 servers blades Cisco UCS B200M3

  Firmware version: 2.2 (3 c)

  Interconnection of fabric are set directly on the fabric of interconnection 6296up and ports are configured as storage FC ports

  My question is this:

  Can I zoning fabric interconnect with FAS3220 and FAS2240?

  May I present LUNS of FAS2240 and FAS3220 at the same time on the same startup strategy, same strategy of startup san and the same organization?

  Thanks for your help

  You can make the zoning on UCS:

  http://www.Cisco.com/c/en/us/support/docs/servers-unified-computing/UCS-...

  If it's really a good practice...

  CF sees WWNNs and WWPN.

  What would be the purpose of pointing both starting the 3220 s 2 and the 2240?

• two or more frame with fabric in cluster interconnect

  Hi experts

  I just want to confirm the proper way to connect two or more UCS 5108 chassis with a pair of interconnect fabric 6120xp clustered.

  I though that they should connect like this: the tissue Extender first of all chassis must connect on interconnection fabric A and second all chassis tissue Expander connects to the fabric interconnect b. as on the diagram here: https://supportforums.cisco.com/docs/DOC-5949

  However, I saw someone did: 1 chassis connects as I went through although above but for the 2 chassis, fabric Extender 1 connect to the B interconnection fabric and fabric fabric Extender 2 interconnect a. -

  can someone tell me which is appropriate/best practices? no reason the second chassis enjambment?

  Thank you

  Ming

  Ming, you must fix this implementation.  It's like ghostbusters: do not exceed the streams!

  Always systematically cable chassis fex to interconnection fabric.  Most common installation:

  Interconnection of fabric (primary) is connected to the hand (looking for the rear of the chassis) left FEX

  Fabric interconnect (subordinate) B is hand-bound straight FEX.

  Why?  You will get errors (can't be bothered them up!) and if you move a Service profile to a chassis to another chassis that is wired differently, it will not work (CNAs wired to different financial institutions, where the living are active / inactive = bad).

  PS - to change this, if I were you, I would like to turn off the entire system - fabric interconnects, chassis, everything.  then re-cable.  then turn on the fabric of the interconnections and chassis and check configurations are ok.

• Can not pass traffic with label of vmware virtual switch fabric 10 GB

  Hello

  I need to understand how to move traffic labeled VMware VST to these virtual Fabric switch. Blades IBM HS22 connecting internally to the virtual switch between ports 1 to 14. I use 2 external ports (17-18), one connects to the Netgear switch and another to the other switch to virtual fabric. Did the same on the other virtual fabric switch. My Synology rackstation is configured with LUN iscsi that connect to the Netgear switch and I would like to connect my HS 22 rackstation blades. My main concern is that I can't ping the IP of netgear on the same interface vlan. I can ping my Synology diskstation to the netgear which are in the same vlan. NETGEAR and BNT switches are connected by cables DAC SFP +.

  

  Even VLAN is also configured on Netgear switch. The default pvid is set as 1 on all interfaces, can I disable this? Do I need to use tagpvid-penetration on all interfaces.

  SH run

  Current configuration:
  !
  version "7.8.7.
  switch type "IBM Networking OS virtual fabric 10 Gb Switch Module for IBM BladeCenter"
  iscli-new
  !
  timezone system 295
  ! Europe/Denmark
  Advanced System
  !

  SNMP-name of the server "BNT01".
  !
  hostname "BNT01".
  !
  !
  enable access userbbi
  !
  INT1 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT2 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT3 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT4 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT5 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT6 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT7 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !

  INT8 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT9 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT10 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT11 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !

  INT12 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  INT13 interface port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  interface INT14 port
  switchport trunk allowed vlan 1, 16-50, 3998-4000, 4095
  output
  !
  EXT1 interface port
  switchport mode trunk
  switchport trunk allowed vlan 1, 16-50, 3998-4000
  output
  !
  EXT2 interface port
  switchport mode trunk
  switchport trunk allowed vlan 1, 16-50, 3998-4000
  output
  !

  !
  VLAN 1
  the name "Default".
  !
  VLAN 16
  name "VLAN16".
  !
  VLAN 17
  name "VLAN17".
  !
  VLAN 18
  name "VLAN18".
  !
  VLAN 19
  name "VLAN19".
  !
  VLAN 20
  name "VLAN20.
  !

  .

  .

  .

  .

  .

  .

  ..

  VLAN 46
  name "VLAN46".
  !
  VLAN 47
  name "VLAN47".
  !
  VLAN 48
  name "VLAN48".
  !
  VLAN 49
  name "VLAN49".
  !
  VLAN 50
  name "VLAN50".
  !
  VLAN 3998
  name "iscsi".
  !
  VLAN 3999
  name "vmotion".
  !
  VLAN 4000
  name "mgmt".
  !
  !
  !
  spanning tree mst configuration
  lethosting-name "region1".
  revision 2
  output
  !
  spanning tree mst mode
  !
  spanning tree mst configuration
  example of 1 vlan 16-50
  instance 2 vlan 3997,4000
  example 3 vlan 3998-3999
  output

  The configuration is for Teddy. I donno what I'm missing here? Any ideas would be very appreciated.

  Yes. Finally managed to do work. Tagged traffic now connects blades with ESXI 5.5 U2 to the Synology rackstation.

  It was the same thing we had. ESXi 6.0 is not supported by this adapter emulex. Also ESXi 5.5 does not I think with the iSCSI driver. So I have updated drivers using esxcli.

  VMware

  Updated network driver

  / tmp # software esxcli vib install v - /tmp/elxnet-10.0.575.9-1OEM.550.0.0.1331820.x86_64.vib
  Result of the installation
  Message: The update completed successfully, but the system must be restarted for the changes to be effective.
  Restart required: true
  VIBs installed: Emulex_bootbank_elxnet_10.0.575.9 - 1OEM.550.0.0.1331820
  VIBs removed: VMware_bootbank_elxnet_10.0.100.0v - 1vmw.550.0.0.1331820
  VIBs ignored:

  ISCSI driver update

  / tmp # software esxcli vib install v - /tmp/scsi-be2iscsi-4.6.261.0-1OEM.550.0.0.1198611.x86_64.vib
  Result of the installation
  Message: The update completed successfully, but the system must be restarted for the changes to be effective.
  Restart required: true
  VIBs installed: Emulex_bootbank_scsi - be2iscsi_4.6.261.0 - 1OEM.550.0.0.1198611
  VIBs deleted:
  VIBs ignored:
  / tmp # software esxcli vib install v - /tmp/ima-be2iscsi-4.6.261.0-1OEM.550.0.0.1198611.i386.vib
  Result of the installation
  Message: The update completed successfully, but the system must be restarted for the changes to be effective.
  Restart required: true
  VIBs installed: Emulex_bootbank_ima - be2iscsi_4.6.261.0 - 1OEM.550.0.0.1198611
  VIBs deleted:
  VIBs ignored:

  esxcli system set to true EI maintenanceMode
  esxcli system shutdown reboot - r = driverupdate d = 10

  After that, I created iSCSI vmkernel ports with the grouping. Ping has started working and now I can connect to the storage

• question from newb on 6100 fabrics

  Hey,.

  What is the difference between the 6100's and nexus switches?

  I look at vids on UCS and it seems that the networking and the fabric are on the 6100 6100 modules are s of the nexus switch? They are completely serperate devices this site to the South the Nxs switches?

  Can someone give a brief explanation?

  Brent, the interconnections of fabric UCS (FI) are the point of grouping all the chassis connection blade & rack UCS servers.  In addition, the FI to run the software of management of the UCS.  FIS are a separate device.  They have just the same physical appearance as the nexus 5K, but they are distinct & unique devices.

  UCS can uplink in your Distribution, or even of the basic network switches.  It could very well be a pair of Nexus, Nexus 7 K, Cat6K 5Ks or any which non-Cisco switch.

  Here's how common a UCS deployment would look like & sit in a network running continuous.

  

  Here is an another digram showing the UCS uplink for both SAN and LAN Ethernet fabrics.

  

  HTH,

  Robert

• 6120 UCS connect to the Brocade 300 FC switch

  We buy a UCS with 6120's and will connect the CF modules to our 8 GB Brocade 300 switches.  My questions are:

  1 are there problems with the connection of the 6120's to the 300?

  2. can I have two separate fabrics?

  3. are there potential loop or latency problems as I go to my EMC CX4 of 6120's by the 300?

  Basically, any comment is welcome :-)

  Thank you!!

  1. Brocade switches will work very well connected to the 6120. The main requirement is that they support NPIV.
  2. Yes you can, and indeed, what would be the recommended path. Fabric Interconnect will connect to A Brocade, while the fabric interconnect B connects to the Brocade B. The interconnections of fabric or the brocades are interconnected, leaving you with two different fabrics. In each UCS Manager Service profile, you can then set up a vHBA for each fabric.
  3. No loops exist, although the fabric of the interconnections will be set up a small amount of latency, as shown in this link. Tiny enough to be negligible if and as the link explains the fabric is lossless.

  http://www.Cisco.com/en/us/prod/collateral/ps10265/ps10276/data_sheet_c78-524724.html