Balancing of LUN paths on your ESX host

Similar Questions

• How to balance IO SAN controller between my ESX host?

  Hello good community.

  Question about SAN controller i/o balancing between my ESX hosts.

  I've identified the controller A and B controller on my SAN, HP EVA 6100.

  The controller has 50001FE1 500E1208

  Controller B 50001FE1 500E120C

  But when I go to "Storage adapters", then "Manage paths" by vmhba and I'm lost.  Policy of LB is fixed.    Then I see:

  vmhba1:0:6 50:00:1f:e1:50:0e:12:08 * status: active, preferred

  vmhba1:1:6 50:00:1f:e1:50:0e:12:0 c

  vmhba2:0:6 50:00:1f:e1:50:0e:12:09

  vmhba2:1:6 50:00:1f:e1:50:0e:12:0 d

  If the first two on my controllers, don't know about the bottom two WWWN however, can someone who knows HP EVA help here.

  Also, I am very confused regarding mirrored ports and the need to be cautious when balancing the way between the controller.

  Which would be great if someone could give me a few white read books on it, the same and also fair knowledge base articles explain how to approach the above and examine the virtual machines as well.

  Not too big so... just small forum post I raised here... It will be useful someone help to get a dialogue going.

  What I do with my EVA, is having different LUNs (Datastores) manually selected a path preferred in command mode.  For example LUN 1,3,5,7 are on path then that lun 2,4,6,8 are set on Channel B.  This means that all your LUNS are also responsible.  On the ESX side each of my machines has 2 HBAs each with 2 paths that it looks like your Setup.  Here's what I think looks like load-balanced:

  ESX01-

  vmhba1:0:6 50:00:1f:e1:50:0e:12:08 - LUN 1.2

  vmhba1:1:6 50:00:1f:e1:50:0e:12:0 c - LUN 3.4

  vmhba2:0:6 50:00:1f:e1:50:0e:12:09 - MON 5, 6

  vmhba2:1:6 50:00:1f:e1:50:0e:12:0 d - MON 7, 8

  ESX02-

  vmhba1:0:6 50:00:1f:e1:50:0e:12:08 - MON 5, 6

  vmhba1:1:6 50:00:1f:e1:50:0e:12:0 c - MON 7, 8

  vmhba2:0:6 50:00:1f:e1:50:0e:12:09 - LUN 1.2

  vmhba2:1:6 50:00:1f:e1:50:0e:12:0 d - LUN 3.4

  It might not even be necessary have a configuration of blancing of load complicated if you produce not enough IO to saturate your current configuration.  You can just keep it like this:

  ESX01-

  vmhba1:0:6 50:00:1f:e1:50:0e:12:08 - LUN 1.2

  vmhba1:1:6 50:00:1f:e1:50:0e:12:0 c - LUN 3.4

  vmhba2:0:6 50:00:1f:e1:50:0e:12:09 - MON 5, 6

  vmhba2:1:6 50:00:1f:e1:50:0e:12:0 d - MON 7, 8

  ESX02-

  vmhba1:0:6 50:00:1f:e1:50:0e:12:08 - LUN 1.2

  vmhba1:1:6 50:00:1f:e1:50:0e:12:0 c - LUN 3.4

  vmhba2:0:6 50:00:1f:e1:50:0e:12:09 - MON 5, 6

  vmhba2:1:6 50:00:1f:e1:50:0e:12:0 d - MON 7, 8

  Hope this helps

• How LUNS presented to an ESX host

  Dear team,

  How to confirm weather LUN is presented to an ESX host?

  We asked our team to SAN map a 900 GB LUN to one of the esx host, these people have done enough for their purpose (e.g. zoning part etc.) and confirm the same thing for us. We rescan the HBA, still, we are not able to see the same thing. SAN team asked us to reboot and check, but it's prodcurion esx box we will not be able to restart, so I want to confirm end VMware, is it possible to confirm weather esx is able to perceive the lun or the requested LUN is delivered correctly to an esx host, needs your help to understand this issue.

  concerning

  Mr. VMware

  If you need create a vmfs on this lun data store, then you can simply follow the add storage wizard via - see this link for more information/screenshots: http://www.vmadmin.co.uk/vmware/35-esxserver/109-esxvmfsdatastore

• Manually load balancing / setting of the LUN paths.

  I am wanting to manually load balance my Lun, there are 4 paths, and I want to cycle each round logic unit through the other way and back loop number. Am on vSphere 4 Update 1.

  Browsing the web I found these two articles provide the code I'm after but the two error for me.

  http://vmjunkie.WordPress.com/2009/01/29/balancing-LUN-paths-on-your-ESX-hosts-with-PowerShell/

  http://doitsmarter.blogspot.com/

  -

  ---

  1. Check with the user

  $vc = Read-Host "Please enter the name of the vCenter server:

  $uname = Read-Host "Please enter the user name to connect to vCenter server:

  $upass = Read-Host "Please enter the password that is associated:

  $clusterName = Read-Host "Please enter the name of the Cluster:

  Write-Host "Connection to the server vCenter $vc, please wait." - black BackgroundColor - ForegroundColor white

  SE connect-VIserver $vc - user $uname - password $upass | Out-Null

  ###

  1. Recover clustered ESX hosts

  $VMHosts = get-Cluster $clusterName | Get-VMHost

  1. Run through this loop for each host in the cluster

  foreach ($VMHost to $VMHosts)

  {

  $luns = @($VMHost | get-scsilun-luntype disque | where-object {$_.) {ConsoleDeviceName-like ' / vmfs/devices/drives/vml * "} | Sort-Object CanonicalName)

  $firstLUNPaths = get-ScsiLunPath $luns [0]

  $numPaths = $firstLUNPaths.Length

  $count = 0

  foreach ($lun to $luns)

  {

  If ($count - ge $numPaths)

  {

  $count = 0

  }

  $paths = get-ScsiLunPath - ScsiLun $lun

  $lun | Game-ScsiLun - MultipathPolicy fixed PreferredPath - $paths [$count]

  $count += 1

  1. Sleep 30 seconds to avoid saturation of the picture.

  Start-Sleep - 30 seconds

  }

  }

  -

  ---

  Initially, I found the script stumbled upon the $luns [0] which means the value is null or empty. I replaced vml with naa. Now, I got "one or several mandetory missing settings SCSI LUN". An error with the syntax, but I'm not able to establish that, I guess.

  Now, if I run certain commands on the command line I find I'm out usable. But by channeling the same row in a table or the variable, then the Write-Host of the content, I see "VMware.VIMAutomation.ViCore.Impl.V1" (and so on).

  Judging by the dates of the blog posts that I am of the opinion the latest PowerCLI (v4 build 264274) can return the information in a different way, I'm wrong with my assumption?

  Can you guess what will be my next question? How could I do that now?

  Any help is appreciated and rewarded (in points not beer).

  Thank you

  Darren.

  The "vml", "naa"... depends on the type of storage are your LUNs.

  The problem with this statement is that you must specify the parameter - Scsilun.

  This can be a positional parameter (without - Scsilun), but then it has to come in position 2.

  See the page Get-ScsiLunPath .

  Try the attached script.

  I changed this line and it works for me.

  ____________

  Blog: LucD notes

  Twitter: lucd22

• Presentation of ID host ESX host LUN to the VCB Proxy Server

  I have a typical scenario where I have 3 ESX servers with 3 x 800 GB LUNS in a storage group separate on a CX 310 SAN box. The host LUN., of these 3 LUN Ids are 0, 1 and 2.

  I have a specific host that starts from SAN, from the same box of CX 310 SAN. At this point of time, I have no choice to use this server as a Server Proxy VCB. As far as my knowledge goes, LUNS of the ESX servers must be made visible to VCB Proxy Server, as they are presented to the ESX servers.

  My question here is, as the designated Proxy VCB server is booting from the SAN, he already has a host LUN 0 to Id the logic unit number this is the start of. The Id to host ONE of the first 800 GB LUN in the ESX Server storage group is also Host Id 0. Now, if I try to make the LUNS in storage group visible on the VCB Proxy Server ESX, is there not a shock of the VCB Proxy 0 host Id and Id host ESX Lun 0? I'm confused, I don't have a test envirinment where I could try this before I could actually go forward with any solution.

  My intention is to perform a backup outside LAN of the VMDK.

  Any help is that this respect is very much appreciated. Thanks in advance.

  (1) Yes, you can have different host for VCB then for your ESX host LUN ID

  (2) Yes, you can use the identifiers of naa, it will automatically use the identifiers of naa, no need to configure anything at all.

  Duncan

  VMware communities user moderator

  -

• Naming of the ESX hosts with the underscore. Best practices?

  I was curious as to what that the general eager to name the ESX hosts with underscores "_". I found an old post in 2006 related to it:

  http://communities.vmware.com/thread/64885

  In the thread, it is usually frowend against because she not compling with DNS internet naming standards. The underscore character is not recognized by the Web servers.

  He has been wondering if there would be no advantage to include similar underscore characters with no routable ip subnets, 192.168.x, 10, etc...

  If you have designed it correctly you already have your hosts on a subnet that is not routable, so you would get no real benefit.

  Beyond this thought experiment, there are problems with your esx host with ESX, vSphere, vCenter naming universe underscores?

  Just an info

  All my guests have underscore in them. I had to replace all of them with a dashboard because ESX 4.1 will not accept underscore after upgrade to the console.

• What is the STIG RSS script and how implement us it with ESX hosts?

  I have heard of STIG script but do not know what he done to the ESX host and what is it and is it necessary? Any ideas would be excellence.

  It is one of the specially designed UNIX script for the Department of defense uses to run against your ESX hosts to any of the configurations and security issues and will generate a report of vulnerabilities.  You can download them directly from the site Web of STIG and run the script to see how vulnerable are your hosts.  Here's a long discussion on this topic and the updated release of the script as it goes.

  http://communities.VMware.com//thread/145435?TSTART=15

  If you found this information useful, please consider awarding points to 'Correct' or 'useful '. Thank you!!!

  Kind regards

  Stefan Nguyen

  iGeek Systems Inc.

  VMware, Citrix, Microsoft Consultant

• Virtual ESX host?

  First of all I must appologise, I tried searching, but my browser gives me a 404 error whenever I press the search.  So if a topic of this nature already exists, please let me know.

  Now to the question.

  I'm looking using Vizioncore vReplicator software, recommends that you have your disaster recovery ESX hosts in a separate group.  Ultimately, this means that your ESX hosts in this cluster will be doing nothing but waiting for a situation of DR.

  Is it not possible to install ESX and then create 2 new virtual machines:

  (1) a production ESX host

  Host 2) A DR ESX

  This way, just host would not be "wasted material" as production VM would BE able to help the production cluster workload, while DR VM would practically nothing (no pun intended hahah)... in a DR situation, the prod THAT VM would eventually not having any VMs to manage (because it would be likely that my just blown SAN) and DR site (with SAN backup) then resume the treatment.

  Any thoughts?  Can I install VMware ESX as a virtual machine in ESX?

  No you can not install ESX in the virtual machine under ESX, but you can under my computer - but ESX running in a virtual environment is not the way forward for DR performance would be catastrophic during a power outage in my humble OPINION

  If you find this or any other answer useful please consider awarding points marking the answer useful or appropriate

• Change the path information for ESX ESXi hosts to access

  Hello all! I hope you guys can help me.

  I have about 47 hosts that I need to change it Round Robin storage path information. If I make the change in vCenter and restart the host it again changing fixed. So, I thought that I should change it to the level of the host through esxicli.

  I have about twice as many LUNS of the Hitachi SAN. I'd like to see if it is possible to make the change in all areas, in some way, or I need to make the change by host? Also, is there anything that I need to change sides San so that it works or not.

  I saw this VMware KB: modification of path information for ESX/ESXi hosts to access and assume that it is the best plan? But did not know of all put into custody, if necessary, that everyone ran to when doing this?

  You will need to make the change by the host and the KB you'll find is the right method.

• 6 zoning LUNs esx host

  Hello

  We have 6 esx host with 2 switches and 300 fc HBA each.2 Brocade on hus 110 with raid 5 7 + 1. !

  the scenario, I don't need a store of data for virtual machines or maybe another 2.

  in the storage Bay, I would add the LUNS in a port and then i will be the area each esx with this port, combining the fc switches 2?

  or should I make another raid group or LUN to the other port of storage?

  y at - it a pattern?

  You will need to create 4 areas by host, two on each FC switch, so a total of 24 areas for your 6 hosts with storage.

  Edit: has a quick and dirty "drawing." Take a look at the attached jpg. When you start storage, there are 4 possible ways to reach each server, which are your necessary areas.

  Concerning

• Get the LUN of the ESX host Info

  Hello:

  Is there a way to extract the info from the ESX host LUN?  I need (s) Lun sees number/ID to match the storage admin ID...

  Thank you very much!

  Of course, the easiest way is this

  $report = foreach($esx in Get-VMHost){
      foreach($hbaKey in ($esx.Extensiondata.Config.StorageDevice.ScsiTopology.Adapter | where {$_.Adapter -like "*FibreChannelHba*"})){
          if($hbaKey.Target){
              foreach($tgtKey in $hbaKey.Target){
                  foreach($lunKey in $tgtKey.Lun){
                      $hba = $esx.Extensiondata.Config.StorageDevice.HostBusAdapter | where {$_.Key -eq $hbaKey.Adapter}
                      $lun = $esx.Extensiondata.Config.StorageDevice.ScsiLun | where {$_.Key -eq $lunKey.ScsiLun}
                      Select-Object -InputObject $lun -Property @{N="Host";E={$esx.Name}},
                          @{N="HBA";E={$hba.Device}},
                          CanonicalName,DisplayName,@{N="LUN";E={$lunKey.Lun}}
                  }
              }
          }
      }
  }
  $report | Export-Csv "C:\lun-report.csv" -NoTypeInformation -UseCulture
  

• Make a single LUN can be used by multiple ESX hosts

  Is it possible for a single LUN to be made usable by multiple ESX hosts at the same time?  For example, if you have an image or used by a vSphere environment requiring other vSphere environments in the data center access, you can put the image on a storage place where completely separate ESX hosts vSphere environments completely separate can see the same lun and clone the image or necessary?

  If the logic unit number is locked by accessed by an ESX host will be lock expires when the ESX host is not an another ESX host in another environment vSphere allowing access to the logical unit number, or is the lock, as it is usable by a single ESX host, until the storage is reconfigured to make it usable by a different host?

  Is there a better solution to this scenario?

  Thank you

  You can deploy both at the same time.  Locking occurs only for changes to metadata and only for the shooting of subsecond.  It is not required for the entire deployment process.

  -Matt

  VCP, VCDX #52, Unix Geek, Nerd of storage

• ESX host managed by VirtualCenter servers all see the same SAN LUNS / data warehouses

  Scenario of

  I have two separate deployments of VirtualCenter (in the same physical location) VCA and VCB with each deployment management 5 x ESX hosts (10 in total).

  All 10 hosts are allowed to look at the same SAN LUNS / data warehouses?

  Or I have to isolate five VCA hosts and five guests to their own SAN LUNS VCB / data warehouses?

  Box293 wrote:

  Basically, we have our direct environment and we are on the beta program for VI4.

  I want to turn off a virtual machine in the direct environment and then pull it out of the inventory.

  Then I want to go the VI4 Beta and browse the data store and add this virtual machine to inventory.

  These two environments may see the same LUNS / data warehouses.

  I did and it works however I want to make sure that it is an approved practice.

  You can do that normally without any problems. Even if the virtual machine is in a different datacenter inventory and there Powred-Off. You can map the logic unit number on the VI4 and add to the inventory of the VI4 another host.

  Recently, I did this practice to migrate from host with virtual machines in a different subnet to the host who knows the same setting SAN LUN in the other subnet.

  Best regards

  Hussain Al Sayed

• VI3 vm LUN at the recommendation of hosts vsphere 4.1

  Hello

  "Note: virtual machines created in an environment of ESX 3.x will work perfectly on ESX 3.x and 4.x hosts." However, when you upgrade the virtual machine virtual material it can only run on ESX 4.x. If you have a mixed environment of hosts 3.x and 4.x, do not improve the virtual material until all ESX hosts have been improved in 4.x. "If the virtual hardware is upgraded before upgrading your ESX 3.x host, then your VMs can only lie on your hosts 4.x.

  I'm looking for the best way to get the VM manged VI3.5 platform, servers, and storage in vsphere 4.1 vsphere Platform of servers and storage.

  What that means I can the current map VI 3.5 to a newly configured LUNS ESXi 4.1 vsphere servers and re platforrm inventory the VM from the installed hosts recently updated?

  Can I map the new LUN storage system to the current VI3, 5 and use svmotion.pl to move them from the virtual computer to new LUNS, map the new LUNS to the new servers and re inventory them?

  What that means I can the current map VI 3.5 to a newly configured LUNS ESXi 4.1 vsphere servers and re platforrm inventory the VM from the installed hosts recently updated?

  Yes - there is a minimum change in VI 3.5 and ESXi 4.1 VMFS file system

  Can I map the new LUN storage system to the current VI3, 5 and use svmotion.pl to move them from the virtual computer to new LUNS, map the new LUNS to the new servers and re inventory them?

  You can it is not necessary

  Maybe support - upgrade to VMware vSphere 4 http://www.vmware.com/pdf/vsphere4/r41/vsp_41_upgrade_guide.pdf Guide

• LUN access is the only host

  Hello

  I have 3 Dell MD3000i iSCSI connected server blade, the three guests can see/access the same logical unit number but when I create a virtual machine under its unique host access by saying is that correct?

  This is probably because you are connected directly to the ESX host. Because the hosts are not managed by vCenter (or if they are, you are not connected to it) your host thinks it's the only host touch this LUN.  Looking at this capture, screen, I would say that it is correct.

  • Kyle