VSA on material different hosts in a cluster

Similar Questions

• Mix the virtual switch different type in a Cluster and a data center.

  Can I mix standard virtual switch and a virtual switch distributed in different hosts to a Cluster/DataCenter? Can I vMotion virtual machine to a host of switch distributed to a host of standard switch and vice versa? Let's assume that the hosts have the same port group name (but have virtual swaitch different type), in the same data center, and have the same subnet IP vMotion.

  You can mix standard switches and distributed, that's what we call hybrid architecture... but to be able to migrate virtual machines between virtual switches, you must the vSphere 6 and again there are some limitations, like not be able to migrate from VDS vs.

  Have a look here for more details on the cross switch vMotion: http://www.vladan.fr/vmotion-enhancements-vsphere-6-0/

• MSSQL server VM cluster should be in the same host or different hosts with RDM

  Could someone me on how to place the code SQL cluster s VM with RDMs in ESXi hosts for advice.

  What is the best practice to place the SQL VM s in ESXi hosts.

  Affinity or an anti-affinite...?

  Appreciated your valuable answers.

  Depends entirely on the use case.

  • 2 MS SQL nodes on the same host to see the availability of the software
  • 2 nodes on different hosts to see the availability of the equipment

  I'd say MS Clusters on the same host (at the time of HA and vSMP FT) are redundant VMware features and represent an increase in management fees. The MS cluster on hosts provide something in addition to what VMware vSphere alone can provide!

• segregation of VM on different hosts

  
  I'm new to vmware releatively, but I am interested in the design of fault tolerance and what are the specific virtual servers that ensure on different hosts. For example in small installations if you had 2 DCs on a single host then who is it a single point of failure in the AD services? Outside of DC, what other specific types of servers/server roles warrant you are on different hosts? Or with VMWARE HA do you not need to worry as much as if his is attached to a cluster virtual machine must be picked up by another host if the host fails, so segregation is not so much a concern?

  Hey cb122,

  Welcome to the community and Vmware,

  First good HA is very different then FT.  HA leans on the hosts and makes sure that all your cluster hosts are running and the computer virtual is running.  If a host fails HA reviews cela and turns on all the VMS running on the host failed on another host in the cluster.  Your VM will see this as an unexpected shutdown.  Although short, this may take a few minutes for all your VMS is to start again based on what was running on the failed host.

  Fault tolerance is basically HA on stariods and what it does is eliminate this short stop breakdown or unexecpted on the host failure.  But that comes at a high cost, reduced capacity, there are a number of requirements to work ect ect, look under FT for more information.  FT back more now is your only allowed to have 1vCPU IP protected VM, which kind of shoot it in the foot because most systems requiring such high availability require more vCPU.   However how FT works in summary is this: you select FT on a virtual machine, and then creates a snapshot of the virtual machine and it works on a different host.  Any changes made to the virtual machine are always synchronized lived with his shot.  If the failure of a host that is running your VM, the VM snapshot is then invited to production and immediately resumed all of its features, it generates on an another snapshot on another host in the cluster to its protection.

  Awenser your question each environment is different, but most of the time a cluster of 3 + host Vmware will protect you for most, but in your case I would put a rule of DRS in to divide your virtual DC 2 as if both are on the same host and that host fails, you will lose all AD and probably DNS as most people run these 2 services together.  This could be problmatic to your environment.

  Aside from that, I would be just shared any VM that could not handle all seen down at the same time, if a failure of the host.  A good example of this would be a web server farm.  Say, you have 4 IIS servers in a web farm, you do not want to run all 4 on the single host as if that host fails your website goes down for 2 to 4 minutes until these web servers would start up on a new host.  In this case, you could create a ruleset DRS to keep 2 webservers on a host and 2 on the other.

  I hope this helps.

• No network connection between the virtual machines on different hosts ESXi

  I have several 5.0 ESXi hosts in a cluster managed by vCenter Server 5 and I have network problems:

  1. [Host A] running a virtual machine can ping any other computer virtual running on the [host A] great!
  2. A virtual machine running on host [B] can ping any other virtual machine running on [B] home - very well!
  3. [host A] (from the ESXi console) can ping any virtual machine running on [B] home - even better!
  4. [B] home (from the ESXi console) can ping any virtual machine running [host A] - beautiful set!
  5. A virtual machine running [host A] can not ping any virtual computer running on host [B] and vice versa - bad, very bad!

  So basically the problem is that the virtual machines can communicate on hosts. It is a major problem that I need to fix, for obvious reasons.

  Networking on all hosts is configured exactly the same way: a unique and standard vSwitch with a pair of gigabit network cards grouped, vmkernels configured for the vMotion, NFS and network management and a group of ports configured for the VIRTUAL local area network used by the VMs. (vMotion works fine on hosts, storage vMotion works very well too.) All virtual machines are servers Windows (2003 and 2008).

  Reason #3 and #4 above, my assumption is that #5 must be the result of an ESXi/vCenter configuration issue and not a problem with the switches of material upstream. Is this a reasonable assumption?

  Whatever it is, can anyone offer suggestions on how can I fix? It is quite annoying as I'm pretty sure that these hosts are configured in the same manner as those of our production cluster, which does not at all these problems!

  Thanks in advance for any help you guys can provide.

  To be honest, it looks like a problem with your physical network-

  How is the physical network configured - are NIC connects it to the same physical switches? Are the virtual machines on the same subnet? If this is not the case, they are able to ping the gateway sbnet? Can you piing host IP address from the virtual machines?

  I also moved it to a forum more approiate.

• How to create disk shared across virtual machines on different hosts

  Hello

  We implement Oracle RAC on two virtual machines, one on each of the two armies for redundancy. For our installation:

  • The two ESXi hosts two SAN access a logical unit number.
  • Two virtual machines on the hosts must have access to a single virtual disk on this LUN (or unit number logic itself) at the same time.

  We tried the following:

  • Creation of a data store associated with the LUN, and then creating a new hard drive on each virtual machine using the same virtual disk on the data store. This operation failed because as soon as a virtual computer runs, the other virtual computer cannot run since the first has a lock on the file from the disk.
  • Creation of a new hard drive on the virtual machine through RDM. This operation failed because only the first virtual machine has this option of disc. Once the ROW has been assigned to the LUN, the other virtual computer cannot create its own RDM for the LUN.

  I'm learning just VMware so I can miss a fundamental method to accomplish this, or perhaps, it is not possible at all. You know a way to create the installer I described?

  Welcome to the community,

  I can not specifically help with the Oracle RAC, however, I would recommend that you take a look at MSCS documentation for how to configure shared disks. There is a difference in the way in which you have installed the VM, for example all VMs on an ESXi host or on different hosts. Please keep in mind that MSCS is a cluster without sharing, which means that only one node at a time has access to a shared virtual disc/RDM.

  André

  PS: I forgot to add the link: http://kb.vmware.com/kb/1004617

• Maybe to lock a virtual computer to a host in a cluster?

  Hello!

  We have a virtual machine that will require a special material (to test it). We step only HW in an ESX host in our cluster. Is it possible to lock the only VM in this specific host?

  Concerning

  H

  With vSphere 4.1 DRS host-affinity has been added (see http://www.vmware.com/files/pdf/techpaper/VMW-Whats-New-vSphere41-HA.pdf), you can set a 'must' rule to ensure that the virtual machine only works on a specific host.

  André

• How to keep the 2 VMs on different hosts

  I have two critical vm and want to make sure they stay on different hosts at all times.  How can I do?

  We have 5 IBM hosts in a cluster of ESX 3.5.  Thank you.

  create a rule of DRS to "separate virtual machines.

  change the settings of your VMware DRSrulesAdd cluster

• Flight hover text on my site by default, ignoring my HTML mouseover. Two different hosts show the same bug. (Chrome showed same on one host.)

  My hover/mouseover stylesheet correctly texts show when < my custom > .css is local disk (offline), but when online, a default value "CTRL + R improves the quality of this image...". "watch instead, presumably from the browser. Sometimes a test shows my stationary custom text on the FIRST event mousover ONLY, then the default value is used.

        This behavior showed also on Crome before I banished it. (Chrome's default text was identical) Support at my usual webhost has been looking at the problem for two weeks without result. Today I loaded the same code to a different host organization (under http://waggledance.us) and I'm seeing the same default behavior, including an occurrence of the "first time okay" observation.
        I note there is a default "Main.css" on that second server, which makes me ask whether a default .css is getting slammed on my user after <my-custom>.css serves (on both hosts??).
  

  This isn't a problem on your web server, it is almost certainly a feature of your internet service. Please see this recent thread for a few possible approaches: How can I get rid of the tooltip improve boring Images? Can solve you?

• Why is no longer a faster network between two different hosts?

  Hello community!

  I need help in this mystery.

  I have three computers different virtual, all of them with e1000 (also I tried all VMs with vmxnet3) and the latest vmware tools.

  Two of this mV, are in the same physical host (ESXi 5.5.0 2068190). Call for example: VM - 1A & VM - 2A

  Another virtual machine, is in another physical host (ESXi 5.5.0 2068190). Call VM-1 b

  Both ESXi hosts are connected with two cards 10 Gbit network with load Balancing: "From hash IP route" and detection: "link status only.

  Traffic Shaping is disabled in all Ports.

  The environment is very clean, and we have more VMs that can disrupt the test (or inappreciable).

  VM - 1a: 10.12.50.201 - physical host ESX - A

  VM - 2A: 10.12.50.202 - physical host ESX - A

  VM-1 b: 10.12.50.200 - physical host ESX - B

  Why I don't see better network speed reference, when two virtual machines send something two physical host is different than when the two virtual machines are in the same physical host?

  It is not a question of "iperf", I used the other test with CPS and I had the same speed.

  Thank you very much!!!

  PS: The physical host computers are: Dell PowerEdge r.620 - 16-core Xeon - 384GB RAM

  

  Have you tried to move your VM on different hosts to the same host?

  Your names of:

  VM - 1 a host has

  -VM-1 b on Hots B

  ToDo

  Now move VM-1B for A Host and repeat the test (two virtual machines on the same host). What happens now to the speed?

  Check

  -How is the CPU usage during the test we Host-A, if you have 2 virtual machines on this host?

  -What is your ratio of P2V CPU? (overprovisioning?)

  What it could be

  -Host doesn't have enough CPU power to serve the two virtual machines (or over-provision P2V related problem). But for me no sense if I see your config server with 16cores.

  -It is not a host but a related problem of VM (perhaps it depends on the VM - A2). Makes the most sense for me (your test is to check)

  Please share the results

• Share of CPU between ESXi hosts in a cluster?

  Is it possible to configure a cluster with a pool of CPU resources such as all CPUS of all ESXi hosts in the cluster are in the pool and if a virtual machine starts using more CPU that's current host has it can draw another ESXi host CPU to deal with?

  A virtual machine can use CPU only from the host where the virtual machine works... There is no way to allocate CPU of multiple hosts to a single virtual machine.

• Not be able to address the multiple host in a cluster

  Recently, we have improved our vCenter to 5.5 from 5.1.

  After graduation we are not able to address the multiple esxi host in a cluster.

  We also tried several remediation option.

  OK, then you can move guests with limitation of do this one by one, right? If so, this is an expected behavior, since there is a limit which allow a maximum of a host by cluster upgrade, see:

  

  https://www.VMware.com/PDF/vsphere5/R55/vSphere-55-configuration-maximums.PDF

• PowerCLI Script required to identify all clusters in a data center and all hosts within each cluster and calculate it using cpu and ram, avg, min, and max.

  Hi all

  I'm new to powercli and try to create a script to list all clusters in a data center and all hosts in a cluster and calculate min, max and avg cpu usage and ram by the host and cluster. So far, I have tried the below but I can't publish the results of my script.

  $Function = @)
  ForEach ($DataCenter Get-Data Center)
  {
  ForEach ($cluster in ($DataCenter |)) Get - Cluster)) - need help to post the information here and confirm if this is correct or not.
  {
  ForEach ($hosts in ($cluster |)) Get - VMHost))
  {
  ForEach ($vms to ($hosts |)) Get - VM)) - do not know if I called you here functions properly
  {
  $allvms = @)
  $allhosts = @)
  $hosts = get-VMHost
  $vms = get - Vm

  {foreach ($vms in $hosts)
  $hoststat = "" | Select the host name, MemMax, MemAvg, MemMin, CPUMax, CPUAvg, CPUMin
  $hoststat. Host name = $vmHost.name
  
  $statcpu = get-Stat-entity ($vmHost) - start (get-date). AddDays(-30)-Finish (Get-Date) - MaxSamples 10000 - stat cpu.usage.average
  $statmem = get-Stat-entity ($vmHost) - start (get-date). AddDays(-30)-Finish (Get-Date) - MaxSamples 10000 - stat mem.usage.average

  $cpu = $statcpu | Measure-object-property value - average - Maximum - Minimum
  $mem = $statmem | Measure-object-property value - average - Maximum - Minimum
  
  $hoststat. CPUMax = $cpu. Maximum
  $hoststat. CPUAvg = $cpu. Average
  $hoststat. CPUMin = $cpu. Minimum
  $hoststat. MemMax = $mem. Maximum
  $hoststat. MemAvg = $mem. Average
  $hoststat. MemMin = $mem. Minimum
  $allhosts += $hoststat
  }
  }
  }
  }
  }

  $Function | Select the host name, MemMax, MemAvg, MemMin, CPUMax, CPUAvg, CPUMin | Export-Csv "c:\Function.csv" - noTypeInformation

  Any help on this is much appreciated.

  [Ordered] casting was introduced in v3 PowerShell.

  For PowerShell v2, you can use

  $vms = get - VM

  $stat = 'cpu.usage.average ','mem.usage.average '.

  $start = (get-Date). AddDays(-31)

  $report = get-Stat-entity $vms - start $start - Stat $stat - ErrorAction SilentlyContinue |

  Group-object - property {$_.} @entity.name} | %{

  $cpu = $_. Group | where {$_.} MetricId - eq "cpu.usage.average"} | Measure-object-property value - average - Maximum - Minimum

  $mem = $_. Group | where {$_.} MetricId - eq "mem.usage.average"} | Measure-object-property value - average - Maximum - Minimum

  New-object PSObject-property @ {}

  Datacenter = Get-Datacenter - VM $_. Group [0]. Entity | Select the name of ExpandProperty-

  Cluster Cluster Get - VM = $_. Group [0]. Entity | Select the name of ExpandProperty-

  VMHost = $_. Group [0]. Entity.Host.Name

  Name = $_. Group [0]. @entity.name

  CpuMin = $cpu. Minimum

  CpuAvg = $cpu. Average

  CpuMax = $cpu. Maximum

  MemMin = $mem. Minimum

  MemAvg = $mem. Average

  MemMax = $mem. Maximum

  }

  }

  $report | Sort-Object-property Datacenter, Cluster, VMHost name |

  Export Csv report.csv - NoTypeInformation - UseCulture

• License question - do need me a VSAN license for all hosts in a cluster?

  I have a 6 host cluster exist using the VSphere Essentials Plus package with a single server VCenter. Can I buy x VSAN 3 licenses, for example to add the storage to the 3 hosts I have install the VSAN license on and access the VSAN of the 3 other hosts in the cluster data store? (Assuming only 1 CPU for each host.)

  Thank you!

  Hi Daniel,.

  Since Virtual SAN is activated at the level of a cluster, all cluster hosts can consume storage since the virtual SAN data store regardless of how many guests contribute to storage. So to answer your question, you need all 6 hosts in your cluster license, even though only 3 of them contributed storage.

  Also note that the licenses are per CPU so in your scenario if each host has 2 CPUs, you need a total of 12 CPU licenses. I hope this helps.

  Chris

• How to force use of the VM a specific host in a cluster?

  I use vSphere vSphere Web Client 5.5 thru 5.0U2.

  When I create a virtual machine, I specify a host in a cluster as computing resources.  However, when I turn on the virtual machine, it uses another host.   How can I configure the virtual machine to use only a specific host in the cluster?

  Use the DRS-> Virtual Machine Options to disable the level of automation for the virtual machine. Who will be actually anchor her to this host. You could also put the VM on local storage or attach too to a resource (DVD-ROM) on this host, but the road of the DRS is the best way to go.