Design of San switch

Similar Questions

• UCS Fiber Channel FI for San switch wiring & config

  My question is about the correct wiring & vSAN configuration by the UCS FI switch San (Cisco MDS).

  Currently my wiring is as follows:

  Fi1 port 25 - SW1 port 3

  Fi1 port 26 - SW2 port 3

  Fi2 port 25 - SW1 port 4

  Fi2 port 26 - SW2 4 port

  On the side of the switch San, our guys of storage configured vlan10 SW1 and SW2 as vlan20. For this reason, I was not able to have a vSAN by FI, I read is recommended, because the VLAN switch cross FIs. Now, when I created vHBAs I have 4 vHBAs per server, 2 by IFS by vlan if I want redundancy with a switch or a loss of FI.

  Question: Is this the correct configuration on the side of the SAN switch? Do I actually create 4 vHBAs per server? I do not have my incorrect wiring & I should just be linking FI1 to SW1 on two ports & FI2 to SW2?

  Help or advice would be appreciated!

  ~ Tara

  Hi Tara,

  There are two ways of wiring of your SAN infrastructure.  It boils down to a design decision. Traditionally the SAN admins prefer to keep their separate/isolated tissue.  In this design THAT MDS-1 only connect to FI - A and MDS-2 would connect only to the FI - B.

  Another option is to connect each MDS at each END.  This gives each FI SAN fabric access and allows to create double VSANS (that exists on each FI) rather than separate them.

  You don't need four vHBAs by UCS Server - some adapters support no more than two in any case.   What you do is create two vHBA by the Service profile, each host will have access both to VSAN 10 and 20.  What you can do is to shift your tissue of vHBA while 1/2 access your Service VSAN 10 via FI - profiles and the other half access VSAN 10 FI - B.  This will help to distribute the load accordingly.

  E.g.

  First 1/2 of SPs

  vHBA0 - VSAN10 - fabric-A

  vHBA1 - VSAN20 - fabric-B

  1/2 second of SPs

  vHBA0 - VSAN20 - fabric-A

  vHBA1 - VSAN10 - fabric-B

  As you can see a bit of a manual balancing can act.  The best scenario would be to isolate (re-cable) to your network SAN fabrics according to the traditional SAN design above.

  Kind regards

  Robert

• FI6248 FC connectivity with MDS SAN switches problem

  Hello

  We have two FI6248 in a cluster must be connected to two MDS SAN switches, using FC ports.

  However, I am not able to configure the ports on the FI.

  There is no "Set up as FC storage Port" option. I don't see what "configure as FCoE storage Port."

  I have the FC SFP (DS-SFP-FC8G - SW--> 8 Gbps Fibre Channel SW SFP +, LC) installed on the two FI6248 on the first 16 ports.

  No idea what is the problem?

  Kind regards

  The 6248UP unified ports.  The ports can be configured as Ethernet Ports or FC ports.  By default, all ports are of type Ethernet.  The ports that you configure is currently an Ethernet port.  You need config of the ports as FC ports.  Ethernet ports will be on the left side of your FI and FC ports will be on the right side.

  The following link explains how to convert a port UP Ethernet CF.  Since it's on your FI and not a plug-in, you have to restart the FI in last step.

  Configuration of the Port Modes for a fabric interconnect 6248

  http://www.Cisco.com/en/us/docs/unified_computing/UCS/SW/GUI/config/Guide/2.0/b_UCSM_GUI_Configuration_Guide_2_0_chapter_0101.html#task_3161F8BE44DA49F6BA9068A5D7594A6C

  Thank you

  Responsible Dan

  Need help with implementation?

  http://www.Cisco.com/go/pdihelpdesk

• UCS connection to SAN switches

  I set up direct UCS in the SAN before, but not San switches to connect.  My understanding of the main differences between the two is as follows:

  (1) with SAN switches, you configure VSAN.  To do this, go to the SAN-> SAN Cloud tab and configure your vSAN ID here.  Then in your vHBA template, configure it to use the proper vSAN ID you configured under cloud of SAN.  Nothing else to do here?

  (2) whereas with direct-attached storage connect FI - A SPA and SPB and FI - B to SPA and SPB, with san switches will FI - A ' switch-1' and FI - B is going to "switch-2'."  Is this correct?  If so, anyone know why it should be that way?

  No problem, tbone-111

  As much as I know, that two vHBAs are supported at this time.  It is best to use 1 from the face and side b 1.

  The vHBAs that are used depend on which WWPN you zoned on your boot LUN.

  For the following example, assume the following group:

  1.) my vHBAs are named vHBA_A1, vHBA_B1, vHBA_A2, vHBA_B2 (A tissue face a and B for b-side)

  (2.) vHBA_A1 is the first HBA on my side, and vHBA_B1 is the first HBA on b-side

  AND

  3.) WWPN of the HBA is zoned properly my boot ports storage processor (SP) of the LUN on the FC switch

  AND

  4.) WWPN of the HBA is hidden correctly on my table LUN

  Those are logical, ANDS, they must ALL be true to prevent headaches later.

  5.) disk Array A0 and B1 SP ports are connected to my A fabric FC switch.  A1 and B0 are connected to my fabric B.

  6.) this example of strategy will start A fabric be the path of MBR

  Best practices - FYI, having at least 2 SAN political Boot (1-A-side, 1 for side b and distribute among your Service profiles).

  In my startup policy, I check the will to apply vHBA naming option and specify my names of vHBAs (sorry I have no screenshots available.  Everything started with a new company and do not yet have access lab).

  My primary SAN parameter can be: the main target: wwpn of A0 and secondary target: B1 wwpn

  My secondary SAN setting can be: the main target: wwpn of B0 and secondary target: A1 wwpn

  It would be a good idea to speak with the Admins of storage while you set up the first two servers to boot on each fabric for all troubleshooting.

  If the zoning and LUN masking are configured correctly, you should see WWPN listed during the boot sequence of UCS server to the point where the starting CF driver is loaded.

  Of course, download TAC Cisco on the line, if you encounter obstacles that you can not go beyond.

  I hope this helps,

  Trevor

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

  If my answers were useful somehow, please note accordingly.  Thank you!

  Post edited by: Trevor Roberts Jr

• Configuration of Nexus 5 k for SAN switching between UCS 6100 s and NetApp storage

  We strive to get our configuration of the UCS environment make start SAN LUN on our NetApp storage. We have a pair of Nexus 5Ks with the company / SAN license on them and the 6-port FC module. What to do to configure Nexus 5000 to operate as the SAN switch between the target storage and environment of the UCS?

  You see the 6120 WWPN on the N5Ks? What is the result of 'show flogi database' and 'show int br '.

  You have mapped the FC ports on the 6120 for the proper VSAN?

  NPIV is enabled?

• Cisco 6248 FI to Brocade 6510 SAN switch shows Link Down

  Hi all

  His new installation of Cisco UCS with Brocade SAN switch 6510.

  I followed the Cisco UCS to Brocade connectivity Guide.

  Even if the switch side so FI link is not coming. No lights are glowing.

  When I checked the switchshow @Brocade 6510, ports are No_Light status display. Where in other ports, it indicates the number of my storage device WWN.

  At the END, when I click on ports CF, its projection on all State - link down.

  Any who are facing this problem. ?

  Am I missing some configuration steps?

  Kind regards

  Praveen

  This should be a trivial or a very complex problem. I'd rather commonplace.

  Brocade is a CF 16 G switch; up to 8 G negotiation seems to work (storage subsystem).

  Maybe you could try to set the speed on the Brocade 8G port fix!

  Or something with the cables is wrong? Send / receive Crusaders?

  Are shortwave of the MMF SFP? cables connection as well?

  Connect cable from port for example. 31-33, 32 to 34 and see if it works?

• Details of port SAN Switch of vSphere or ESXi

  The way I see my network switch name & port details to which my ESXi network cards are connected via CDP.

  With the help of vSphere or ESXi command in any case I can check SAN switch of details in the SAN switch name & ports Details to which my ESXi storage HBA cards are connected?

  They are Cisco MDS 9513 SAN switches, you know a plugin or a way to see who?

  Yes, using the plugin VIEW, take a look: Cisco DCNM rocks | Jeremy Waldrop & #039; s Blog

• activity maintaiance for SAN switch

  Dear team,

  our storage has informed us that these people are planning activity enjoyed in the weekend next, SAN both switches are in redundant mode so no impact of the UPS
  activity on the availability of the San. However ask you to check it out on the side of VMware so all the virtual machines are configured for multipath & are not dependent on the only network SAN
  switch to access the storage.

  all the SAN data store have 4 path (a path is active I/O and rest 3 paths are active. PSP is fixed)

  I just want to confirm if a switch goes down then esxi will automatically route the redundant traffic available SAN switch or do I need manual intervention to route traffic to redundant SAN switch.

  concerning

  Mr. VMware

  
  

  You must confirm if the 4 paths to LUN are in different tissues, cause maybe you feel peckish 4 paths as a single fabric, and if the switch of this fabric must be down you will lose access to the storage system.

  You have two HBA? Each connected to a different switch?

  If you find that your 4 paths in different fabrics and each HBA has access to all the controller storage system processor, you can safety put down one of fabric and ESXi will be managing the paths downwards.

• Paths always down after returning from a SAN switch at the service

  I had a problem where electricity is cut to one of my SAN switches.  Everything continued to work because all the hosts and storage are connected to a second switch SAN.

  Now that powered the SAN switch back on all paths through the SAN switch are still displayed as out of service.  If I reboot a crowd she restores all of its path through two switches SAN so there is nothing wrong with the zoning on the switch.  I see that the ESXi host host bus adapters connected to the SAN switch, even without reboot, when I do a switchshow (using Brocade switches) and I see that the configuration is correct.

  I tried a new analysis of the HBA, and this removes the paths through the HBA that is connected to the switch that loses power!

  I tried to reset the HBA to get to do a reset of the LIP by publishing: -.

  esxcli fc san storage reset - a < hba >

  for example

  esxcli fc san storage reset - a vmhba1

  and tried to change the State of the railways in assets with

  esxcli storage path together - State = active - path < path-runtimename >

  for example

  esxcli storage path together - State == active - path vmhba1:C0:T1:L1

  None of them have helped and I don't want to resort to restart the host, there must be a better way to bring these paths.

  I use ESXi 5.5 build 1474528.  Want to really upgrade to update 1 immediately.

  I have the advanced of that setting Misc.apdhandlingenable set to 1 and Misc.APDTimeout value 140.  The switch would have been off for more than 140 seconds.  I could try to set Misc.APDHandlingEnable to 0 and simulate the failure once again, but I'm not sure I want to do this because I remember issues with ESX(I) 4.x hosts becoming disconnected from vCenter following situations of CCA.

  Any suggestions?

  I followed it down to the driver used by the Fibre Channel adapter.  I had to update 10.0.100.1 - vmw.550.0.0.1331820 10.0.725.203 - 1OEM.550.0.0.1331820.

  The hosts have been using Adapters Emulex LPe16000 (IBM rebadged in an IBM Flex System).  See 5.5 ESXi hosts don't restore not path storage from the SAN switch failure. Computer services Pelicano so all the details.

• 10GB with 1 GB SAN switch?

  Hello

  I was prompted to install the below: -.

  • Ps4210x EQL
  • 2 x Dell r430 (with NIC 1 GB)
  • 2 x N2024 switches (iscsi)

  My concern is that we would use a 10 GB SAN with a 1 GB switch (NIC 1 GB on the server).

  My colleague says the Equallogic can connect to the uplink of 10 GB on the switch ports, but the servers will be able to connect to the server to 1 GB.

  I am curious to know if this is considered to be good practice - I don't like the sound of it!

  Huw-

  The PS4210 have 2 x SFP +, 2 x 10 BaseT G and a single MGMT Port.  The BaseT are compatible down to 1 G. Your PS is connectet by 2 x BaseT by controller where a controller is active and the 2nd. is in standby. If you can read 200 MB/s of MPIO depeding how good work.

  It's not a problem because the current line of the PS does not contain specific patterns of 1 G, as well as most in good old day with the PS6010<-><->6000, 6110 6100. If you took a decision when you purchase the device. You get two options in a box and can change whenever you like.

  Concerning

  Joerg

• Connection to the SAN switch upstream

  I'm having a problem getting a green light on the interconnection 6248 (end-host FC mode) and IBM/Brocade FC switch (speed of the 4GB port currently) upstream.

  Side 6248, I have a 8 G 4 GB SFP and SFP (Cisco DS-SFP-FC8G-SW) in the SAN upstream switch. 6248 attached to the FC switch ports are configured as uplink ports? What other measures are necessary for interconnections to see the storage EMC VNX Bay?

  Thank you

  Before that, I discovered that you need to configure ports on the interconnections as CF and not the ethernet ports by clicking "Configure the Ports unified" on the General tab for each interconnection and adjusting the slider, and then select the ports (right click)... At least that was my problem... lol

  UCS' happy ing...

• Connect FI two directly attach FCoE storage and SAN switches

  Nobody tells me if I can connect FI two directly attach FCoE and SAN storage switches also, as you can see in the diagram of connected network?

  If so, should I FC Switch function Mode? Any concerns should I consider?

  Or the best solution would be appreciated. Thanks in advance.

  The UCS will be in Switch Mode FC, which means that it has a default area.  We cannot make a Zone merge with Brocade because we have no control to set up specific items such as domain and other articles CF IDS that need to be changed when you connect a Cisco FC switch to a Brocade.

  Usually in this situation with a 5 k or a MDS you might be placed in mode of interoperability, which changes the default settings and we will talk to them together - in this case, you can't change these settings, so the only supported configuration is CF Cisco at Cisco.

  See you soon,.

  Greg Scarlett.

• Adding a user role for SAN switches

  I'll try to find the correct location in 3.3 ACS to add the following: roles = "network-admin". We have our SAN with Ganymede switches +. When one user other than admin connections, you get the role of "network operator". The Cisco MDS 9000 Family Troubleshooting Guide, version 3.x doc explains the role if you are using IOS/PIX Radius. Thank you.

  Hi Ed,

  Here is the link,

  http://www.Cisco.com/en/us/docs/storage/san_switches/MDS9000/SW/rel_2_x/San-OS/Configuration/Guide/cradtac.html

  If you are looking for:

  GANYMEDE + custom attributes can be set on a control access (ACS) server for various

  Services (for example, shell). The Cisco MDS 9000 family switches require custom GANYMEDE +.

  attribute of the shell of service to use for the definition of the roles.

  Cisco ACS GANYMEDE +.

  Shell: roles = "network-admin".

  Shell: roles * "network-admin".

  Cisco-av-pair * shell: roles = "network-admin".

  Cisco-av-pair * shell: roles * "network-admin".

  Cisco-av-pair = shell: roles * "network-admin".

  On GBA, if you go to: the Interface configuration, GANYMEDE + (Cisco IOS), check nex to: "display a window for each selected service in which you can enter custom GANYMEDE + attributes.

  Then, go to the configuration of groups and define the role information according to the above attributes.

  Hope that helps

  Kind regards

  ~ JG

• Fibre Channel SAN / switch?

  can I use a 4200 ProCurve switch with optical modules to connect to a 3 host vmware SAN fibre channel?

  or I need to switch dedicated to SAN?

  My SAN is an Eternus DX80 with 2 fiber channel controller.

  Thank you too much.

  ProCurve's network switch.

  Data a switch needs dedicated, you can't mix and match network with data traffic.

• Request for documentation of design of San...

  This has been posted a few years back, but it seems that an adequate response was never given. Is there a documentation of the available industry regarding sizing and planning for the san.

  "3 SAN sizing - VMware has a range that exceeds not 32 VMs per LUN."

  If these are very used then extends over 100 mV per LUN, so virtual machines

  are not widely used. It is a very wide range. When sizing the

  performance, VMware shall provide the info to back up the above,

  that includes what was the LUN RAID level, size of the LUN, disk number by

  the LUN, disk speed, proc I/O maximum, etc. so that we know why, is to be

  has said. Example, I have a SAN with 1.3 TB of space - why can't be scratched,

  say 28 drives in RAID 10 and have 1 large LUN to accommodate 100 strongly

  servers used? What happens if the LUN has been scratched with RAID 4? Simple

  Guideline sponsor LUN sizing based on the number of PIN based

  RAID levels to evaluate the good design of performance for x number of virtual machines

  should not be asking for too much. »

  Thanks in advance.

  That's the thing about sizing of discs - there is no such thing as a recommendation of mat.

  Your sizing LUNS will be more often dependent on the characteristics of I/O you need for the device - IOPS / s, disk throughput, etc.  The number of virtual machines, you can reasonably put on a virtual machine will also be limited by these figures.  What I'm trying to do when my environments of sizing is to determine a PEAK base line (think that request here the worst case scenario - what happens if all VMs apps will go nuts on the disk at the same exact time) IOPS / s (this is usually the biggest bottleneck of disk - physical is a pain here) as the head on an axis can be in one place at the same time), understanding braces how much I need to meet that.  Once I know the number of batteries that my logic unit number has spread throughout, I can watch the overall size of the data store that I'll need.  Keep in mind not only disk sizing here, but also VM snapshots will be whether to use and for how long and also keep in mind the memory allocated to each of the virtual machines to plan for the dimensioning of the .vswp file.  Once I have calibration information, I can watch the split unit size on my storage array, I need to know how much divides satisfy both IO and size needs for the data store, and then create the LUN being mindful of the air RAID.  A rule of thumb that I still use, it's 1 physical pin provides me approximately 150 IOPS / s to work with.  This number is very conservative (and frankly a little old, but I have not looked to drive too deep in the last 3 years).  Also keep in mind that your LUN will share spindles (and so share IOPS / s) with the other LUNs, another activity on the storage Bay must therefore be taken into account during the design process as well.

  For example, if I have an environment with 10 applications, each with a peak Ops ES / s 500, I want finally to a LUN that would support 5000 IOPS / s.  In collaboration with my 150 IOPS / s as a rule, I would try to LUN covering 34 pins.  If each of the VMS has been configured with the virtual disks of 50 GB and 4 GB of RAM each, and I knew that the VMs would not have any how snapshots would be growing more than 10% of the size of the original virtual disk, I plan for 550 GB of disk space virtual/snapshot, 40 GB for the .vswp file and some overhead for the files of newspapers , the config and VMFS metadata, etc.  So I would to, say, a 600GB LUN.  At this point, my table of storage can support a LUN covering 34 pins of 600 GB?  It's a Go 17.6 divide size.  A compromise can do based on what you know of your application performance (peak load will happen all at once, or is the worst overall peak lower load?)  which could lead to a larger size of split and the axes less) or what you know of your storage array (split sizes are available more large and smaller than the calculations show - are you going with a split more small and more spindles, wasting available i/o, or do you take the large size of split with fewer pins) (, with the possibility to you of the IOPS / s starving or wasting valuable disk space).

  With regard to the maximum number of virtual machines that are placed on a specific data store, you would much more than 32 virtual machines on each, in large part because of the way VMFS file locking is maintained.  Each host locks files in the area of metadata for the VMFS data store.  The locks are updated periodically.  These updates and any other activity to update metadata, causes a lock of all SCSI device be maintained on the disk device.  Above 32 VMs highly used, metadata updates can easily keep the LUN in a locked state, preventing the VM disk i/o to occur.

  It is almost impossible to set up a vast document because the recommendations made to your environment are specific to your environment.  Of course, workloads between different stores are similar, but the subtleties of each shop means that any widespread recommendation as "build 500 GB LUN 10 batteries and put 15 VMs on them" would be overkill for 45% of stores to research and would be a disaster of performance for 45%, leaving a small 10% happy environments and a good comfort zone.

  Hope that makes some sense and helps.

  -jk