MSCS CLustering

Similar Questions

• 2012R2 MSCS clustering on ESXi 5.5 build 2718005 issues

  I am trying to set up 2 2012R2 mscs clustering on 2 VMS using RDM

  tried both physical and virtual compatibility mode and keep them coming with cluster validation errors

  anyone with these issues?

  

  Hello

  Try to connect the disk online, diskpart and clean the partitions, recreate the partiotions and rerun the validation tests.

• MSCS Clustering on ESX 3.5 with NFS

  Hello

  Guys I have advised the installation of MSCS cluster using this article in the production environment

  http://TechRepublic.com.com/5208-6230-0.html?forumid=102 & threadID = 220875 & start = 0 & = glad tag; leftCol

  My environment is ESX 3.5 U4, NFS/TCPIP, NetApp. a few questions?

  Is this possible in NFS/TCPIP environment?

  What efforts copies/taken in charge?

  Recommended?

  I understand it is not supported and does not work on NFS.

  Please provide your comments and reflections

  See you soon

  AJ

  The article describes the use of VMware server, not ESX.  It is using disks shared for clustering, and method of use RDM not supported.  You can create a cluster by using NFS, but you would follow the cluster-in-a-box method, using vmdk shared, which is not a method of care or intended to be used in the production.

  -KjB

• Poor with RDM only when added to the CAB MSCS cluster performance

  Hey guys, thanks for taking a peek at this question... I'm looking for some ideas.

  I have 2 virtual machines running Windows 2008R2.  They put MSCS in place and works as a taxi (cluster across boxes).  I use a 7500 VNX, fiber channel and RDM physical drives.  Is on an 8 node ESXI 5.1 v1117900 implemented.  Functionally, everything works fine.

  The problem is that the performance is very poor on only the RDM that have been added to the MSCS cluster.  I can take the same drive, remove the cluster, run IOmeter and it's fast.  I add to the cluster, MSCS, leaving in storage, and it is 1/15th the IOPS performance / s.  Remove the drive of MSCS and it goes back to when you run as usual.

  I tried different controllers SCSI (LSI Logic SAS vs Paravirtual) and it doesn't seem to make a difference.  I have physical MSCS clusters that do not seem to present this kind of performance problem, so I wonder if it was not something goofy with the MSCS on the configuration of virtual machines.

  I've already implemented in this article: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US & cmd = displayKC & externalId = 1016106

  I saw this poor bus rescan the performance issue until I put the RDM perpetually booked, but it was not a problem since I implemented the fix.

  For any help or suggestion is appreciated...

  Thank you

  -dave

  Recently, I upgraded my esxi from 5.0 to 5.1 environment and have a bunch of systems with RDM. I had a problem as well. What I found, was that during the upgrade it changed all the policies of path to the LUN and rdm, it's the turn of robin. This causes a huge performance issue in my environment. I changed all the paths to MRU and it solved the problem.

• MSCS VMs and snapshots

  I have two virtual machines that are part of the cluster, MSCS running on 4.0 u1 on two physical hosts. I use such physical compatibility mode recommended by doctors.

  My question is, we can take snapshots of the virtual machines that are part of MSCS? Or we can do only if they use virtual compatibility mode (all nodes on one box)?

  Allowed himself to take snapshots, let's say I take the cliché of node 1 and then it come back back 20 minutes later, how will the MSCS react to this change and it will confuse the node 2?

  Any input will be greatly appreciated... Thank you!

  It is not recommended to take the snapshot for VM with RDM in physical comatibility mode lunsie.

  it you are planning to build a cluster in a box with disk vmdk shared, you can try to take pictures but, the answer back to the snapshot is dependent on the response of the general application of the change of data.in I wouldn't recomment instant for MSCS clusters.

  Concerning

  Pierre Gustave Toutant

• MSCS and Datacore storage

  Hi all

  I make several facilities of MSCS clusters in ESX 3.5 and 4 boxes that it worked fine. But now, I'm in trouble with a cluster configuration W2008R2 with a Symphony Orchestra of SAN (FC) Datacore storage server.

  I can't find any documentation talking (except a note on virtualized storage).

  I am able to viem and configure RDM on a virtual machine on a host, but this disc is not visible by the second VM (on a different host).

  The virtual machine has been created to support clustering and disk (RDM) is bound to the physical compatibility.

  I hope someone could help me.

  ARM

  CMO

  Once you have created a RDM on a VM all creating a RDM disk options are greyed out so you can add an existing drive.

  This is normal because ESX (i) sees a free uninitialized LUN so no RDM cannot be created.

  Maybe these notes on configuring Microsoft Cluster Services with servers using Fibre Channel or iSCSI storage can help you:

  FTP://support.DataCore.com/PSP/TBS/TB13_MSCS.PDF

  Or check out this installation guide for the implementation of MSCS on vSphere (although I think that it shouldn't be something new for you):

  http://www.VMware.com/PDF/vSphere4/R40/vsp_40_mscs.PDF

  page 25 discusses adding the RDM existing second MSCS cluster node.

• Number maximum official virtual computers Win2k8 supported in a MSCS cluster

  Please can someone point me on the official document or an answer to the following questions:

  • What is the maximum number of Windows 2008 virtual computers based supported in a MSCS cluster (based on vSphere 4.x cluster)

  • more generally, what is the maximum number of Windows 2008 virtual computers based supported in a MSCScluster (based on any cluster ESX/ESXi)

  All the documentation I've seen only speak of 2 nodes (and other sources mention that only 2-node MSCS clusters are supported by VMWARE), but I want a definitive answer to VMWARE or documentation that specifies the maximum number of VMs of Windows 2008 that can be in a MSCS cluster hosted on ESX/ESXi clusters.

  Thank you.

  All the documentation I've seen only speak of 2 nodes (and other sources mention that only 2-node MSCS clusters are supported by VMWARE), but I want a definitive answer to VMWARE or documentation that specifies the maximum number of VMs of Windows 2008 that can be in a MSCS cluster hosted on ESX/ESXi clusters.

  The documentations are correct. Even though you may be able to create a cluster with more than two nodes, the supported limit is 2 knots.

  Take a look at MSCS documents to http://kb.vmware.com/kb/1004617

  literature vSphere 4.1 (page 34):

  Table 6-2. Other requirements and recommendations of Clustering

  ...

  Windows - only two cluster nodes.

  André

• Vsphere-Windows failover cluster replication

  Hello

  Need help to replicate Microsoft cluster at 2 knots.  I know that vSphere replication is compatible with virtual RDM so will change to virtual RDM physica, but the controller SCSI ROW of nodes have buses sharing (Physics) is enabled. How to reproduce these groupings with SCSI bus sharing enabled, since vsphere replication does not support sharing of bus

  Thanks, waiting for some thoughts/work around

  vSphere replication doesn't support MSCS, see: KB VMware: vSphere Replication FAQ

  Is VSphere replication supported MSCS clusters?

  vSphere replication does not work with virtual disks open mode "multi-writer". Virtual machines MSCS cluster are configured with virtual disks open mode "multi-writer", so vSphere replication will not work with MSCS configuration.

• Best practices - addition RDM to the second node (W2K3) MCSC Machine virtual nodes on the physical host computers

  Addition of RDM to the second node (W2K3) MCSC Machine virtual nodes on the physical host computers

  Unable to find another thread on this

  When you add RAW disks to the second node in Virtual Machines across physical hosts in the Cluster / Cluster across boxes,.

  VMware said shared point of the storage drives in the same location as the first node sharing storage disc *.

  -Select use an existing virtual drive...

  -In the drive path, navigate to the location of the disk (quorum) specified for the first node

  -Select the same virtual device node you chose for the first virtual machine shared storage disks, IE SCSI (1:0)...

  In other words to add the RDM to mscs-node2, navigate to/vmfs/volumes/lun1/mscs-node1 / mscs - node1_2.vmdk (mscs-node1_2 - rdmp.vmdk)

  For years we have directly added the ROW the second node specifying RDM disk not existing does not, in general we do directly from the host, not the vCenter, it seems to work fine.

  For what is the safest way, the official method can cause all sorts of problems if you need to cancel the registration of the ROW on the first node (here's where I found no official documentation).

  Delete you or keep the file descriptor? We tried to keep him, but ended up with several mappings to la.vmdk/rdmp.vmdk, so now, this system has disk2_.vmdk / disk2_ - rdmp.vmdk and disk4_.vmdk / disk4_ - rdmp.vmdk pointing to the same RAW.

  What really bothers me is safety, these are very important boxes, I prefer to continue to have the VMDK and rdmp.vmdk in separate data warehouses, and do not have this dependence on the head node

  Your comments please, we are viewing the store only setup of MSCS clusters with lanes separated from RDM and are there risks associated with this?

  * Ref: "Setup for Microsoft Cluster Service - 4.1 and Failover Clustering".

  I realize there was an error in my logic

  When you work with the main node, if there is a requirement to unmap drives ROUGH (move to another cluster vmware, cloning the system etc.)

  Take note of the location of all the rdmp.vmdks

  Remove each rdm disk without deleting

  To add

  Add "Use an existing virtual disk to disk" (yes I know its bad, but once you create the host calling now think its virtual)

  Navigate to the mapping of existing raw device, appearing like a vmdk * and add using the former location of scsi

  Graphic interface hides the descriptor

  A virtual disk has a vmdk and a flat.vmdk

  A CRUDE disk has a vmdk and a rdmp.vmdk (the dish is replaced by vocation)

  A suggestion of one of my companions is to locate the al the ROW in a single data store small, visibility of virtual machines with raw disk went like this

• PowerCLI, Multipathing change: fixed RDM and VMFS of Round Robin

  With the new improvements in vSphere native mulit pathing (NMP) in 5.1, I tried to change my LUN to use repetition alternated according to EMC with my flare VNX practices code.

  However, I also have MSCS clusters that don't support a fixed mode according to the http://kb.vmware.com/kb/1037959.

  Later, I want to get all my VMFS volumes and set them to Round Robin and let all of my RDMs set to fixed.

  I'm hoping to dynamically so that I can leave it with the admin here to run whenever he wants, but I'm fighting to achieve.

  What I've tried so far.

  I can get all my LUNS with these commands.

  $esxcli= Get-EsxCli

  $Allmyluns = Get - esxcli.storage.nmp.device.list () | Where {$_ .device - like 'naa.*'}

  I can get all my RDM with this command

  $RDMS = get - vm | Get-hard drive | where {$_.} SCSIcanonicalname - like 'naa.* ' | Select SCSIcanonicalname

  I can put the method once I have the storage in this way

  $esxcli= Get-EsxCli

  Foreach ($lun in $allmyluns({

  $esxcli.nmp.device.setpolicy($null, $_.Name,"VMW_PSP_RR")

  }

  I have also thought about writing a csv or text file with the first command and trying to remove the entries with the second command but that is a bit past me, and I may be trying to over-complicate things unnecessarily.

  
  

  But what I can’t do is work out how to get a list whenever I want that has all the VMFS LUNS but not any of the RDM LUNS. So I can dynamically feed this into a script which updates VMFS pathing to Round Robin.

  
  

  Any help, input or ideas would be greatly appreciated

  This should help you all VMFS LUNS.

  Get-Datastore | where {$_.Type -eq "VMFS"} | %{  $_.ExtensionData.Info.Vmfs.Extent | %{    $_.DiskName  }}
  

• Replica groups of ' Always on ' SQL Server 2012 - limited to 2 knots on VMWare?

  VMWare KB1037959 provides instructions for Microsoft clustering on VMWare, but does not have SQL Server replica groups cover 2012 "Always on."

  This KB article sets a limit of 2-node MSCS clusters, but also indicates that "the SQL mirror effect is not by VMware to be a clustering solution. VMware supports fully mirroring SQL on vSphere with no specific restrictions. »

  In SQL, 2012 replica groups are a combination of failover Microsoft Cluster Service (MSCS) monitor and SQL database mirroring.  Yo I have to install the 'Fail on Microsoft Cluster' role to implement the SQL replica groups, master didn't have any shared disks.

  I was not able to find guidance on the use of groups of replica SQL 2012 on VMWare.  I hope that it is supported without restriction (as long as you do not use a shared quorum disk), but a strict interpretation of the KB1037959 leads to a limitation of two nodes.

  Can anyone shed more light on this?

  Now the last row of the first table lists:

  Availability of AlwaysOn SQL group

  Yes

  Yes1

  Yes

  Yes

  Even OS/app

  Yes

  Yes

  Yes

  Yes

  N/A

  N/A

  Perhaps this was added recently.

  That answer your question?

• RDM or no RDM

  I'm currently debating with myself ;-) weather or not to continue to use RDM on the file server is a windows 2003 r2 but I want to upgrade to windows 2008 R2 then here's the riddle. The windows guest operating system is vmfs partition on the lun and the 2nd drive is a ntfs partition physical rdm on HP MSA 1500 lun. I need to move the data on disc rdm to another number of logical unit on HP MSA p2000 g2.

  So it would be a wise move/copy the rdm lun configured as a vmfs LUNS or physical/virtual mode data. The size of the lun will not exceed 1.2 TB we have strict file server data storage controls in place. What would be the fastest way to get data two LUNs; If I have the card 2 x rdm LUN physical windows server, the copy rate is about 30 MB/s, tried in virtual machine environment, a few weeks ago and the copy rate was terrible.

  Then it would be wise move/copy the rdm lun configured as a vmfs LUNS or physical/virtual mode data

  Although one of these options works, I prefer the option of virtual disk, need to ROW for example the MSCS clustering. Using a virtual drive reduced the complexity and the number of LUNS and presentations, that you must maintain. You will also be able to run backups of image based in this case.

  With regard to the copied data. Unless manually align you the partition NTFS of Windows 2003, you use currently, you can receive a copy of the data to an aligned partition (WIndows 2008 partitions lined up to 1 MB). I often use Robocopy (included in Windows 2008) to migrate the data. This command line utility is very powerful and allows you even mirror the target with the source data. In this way, you can copy all data on the target in advance and at the time that you spend on the new server it will synchronize only deltas. Robocopy is also capable of maintaining NTFS permissions and owner if necessary.

  André

• Open a virtual machine in a san on 2 different esx hosts

  Hi, I have a Vsphere essentials, 2 Esx hosts and an Iscsi san, I need to connect the virtual machine in the two esx hosts (I know I don't have FT and HA), in order to launch the virtual machine never the first Esx and if first host, does not have the virtual machine on the second esx.

  I tryied to connect the virtual machine to the second host, but the Vsphere client tell me that's not possible, because the virtual machine is already present at the first ESX.

  There is a way to establish this setting?

  For some it may depend on your application, if you can run multiple instances of the same application? If it is a web application? It stores its data in files on the server, or it stores it in a database locally? or remote database?

  -VM on a local storage of the army, would not really recommended if you need so that it starts the VMS on host b quickly.

  -If the two hosts have access to permits it storage even say and what the application can really be run only once.  He could only save it to HostA and if it fails at some point that it will be released the lock, have it on the virtual machine and you will be able to manually register the virtual machine to host b and start the virtual machine.  However, you will not be able to really get ahead.

  -HA out of here, as if a host or power failure shuts on HostA, HostB will feed into the virtual machine down automatically

  -FT - here in the spectrum creates a VM ghost on the secondary host in levels (what happends on HostA happens on host b)

  -Another option to the other end of things has implemented Microsoft MSCS Clustering with iSCSI support I think.  You can run VMS with iSCSI to your San and clustered like this application.  Just note that with iscsi client it uses some CPU to create iscsi packets.

  -Web Cluster, Cluster database.    It doesn't matter if one of the nodes of web servers or the database nodes go down really that the user will be redirected to another node in the cluster so really no need for HA/FT.

  Of course, with some of the here above are expenses and pro / con of and there's a few ways to increase availability.

• SLES + OCFS2 shared physical RDM

  I intend to install a cluster node Linux SLES OCFS2 using a RDM in physical mode 3.

  I did this once before with both VM and set bus SCSI physical Division on the 1st virtual machine, and then for the 2nd server selected the 1st mapping VMDK file. This gave the desired effect to allow the physical RDM to be shared, but meant that vMotion was not possible for one of these two servers.

  What I intend to do this time is to configure all the same VM 3, everyone has the physical RDM attached without bus sharing together on the SCSI bus.  This is possible, I think that, by setting config.vpxd.filter.rdmFilter to false before you add the ROW to each virtual machine. Then the anti-affinites rules will be set to stop 3 VM running on the same host. This should allow the sharing of the RDM OCFS2 and vMotion of virtual machines.

  I think it will work. Can someone tell me if there are problems with it? Is this a supported configuration?

  Thank you

  Neil.

  More than 2 nodes is not supported by VMware. OCFS2 supports more than two nodes. It has been popular with Oracle RAC, until everyone has opted for the DSO. While VMware does exactly what the no-MSCS Clustering support or requirements are. Most of the things are based on the limitations of MSCS clusters in VMware. I am about to implement a number of clusters for a customer with OCFS2, but will be limited to two nodes. We were just will you use FT for our high availability needs, but with the guest operating system as a single point of failure have failed to meet application SLAS.

• VCenter 4.1: cluster it?

  Hello

  Is it possible to Vcenter 4.1 cluster (2 nodes for example)?

  It was ok with VirtualCenter 2.5, under MSCS, but I don't see anything to do this for Vcenter 4.1...

  Thank you

  Concerning

  In our environment, our mandate is HA or not go.  Clustering is always a must... but for some odd reason, this does not applies to vcenter because it is a virtual machine that is running in a Cluster HA.  As for heart rate its good only to have a standby vcenter server and accessible. It is not mirrored DB, which can still be a single point of failure depending on your configuration.

  A few good insigts thereon:

  http://virtualfuture.info/2009/01/VMware-vCenter-SQL-Server-best-practices/

  Official KB in this issue:

  http://KB.VMware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalID=1024051

  VMware, said "Yes you can do it... but.

  «You can choose to protect VMware vCenter Server with third parties including, but not limited to, MSCS (Microsoft Cluster Services) and VCS (Veritas Cluster Services) clustering solutions.» VMware does not certify these third party solutions. VMware will support all of the problems with an environment that uses third-party solutions from VMware vCenter Server downtime protection. However if your problem is considered to be related to the third party, clustering solution, VMware will send to our policy of third-party software. »

  And of course yellow bricks Duncan has something to say about it as well:

  http://www.yellow-bricks.com/2010/07/16/MSCS-clustered-vCenter-Server-4-x-not-supported/

  Now Vmware HA - heartbeat is the method of SUPPORT ONLY...

  See you soon,.

  Chad King

  VCP-410. Server +.

  Twitter: http://twitter.com/cwjking

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