Quantcast
Channel: High Availability (Clustering) forum
Viewing all 4519 articles
Browse latest View live

Creating the DHCP failover relation from remote server

$
0
0

Hi 

 I am trying to establish the failover relation ship between two windows servers (w1 and w2 which are 2012 version) with power shell commands like
Add-DhcpServerv4Failover -ComputerName "dhcpserver.contoso.com" -Name "SFO-SIN-Failover" -PartnerServer 10.0.0.99 -ScopeId 10.10.10.0,10.20.20.0 -LoadBalancePercent 70 -MaxClientLeadTime 2:00:00 -AutoStateTransition $True -StateSwitchInterval 2:00:00

but this command will be executed on one of windows server from an external linux machine(external remote), where winrm being the interface between the linux and windows. This winrm will carry the command from linux and gets executed on w1 at that time it is trying to communicate to another windows server w2 to cross check failover already exists or not but it is unable to communicate w2 stack trace gives me permission denied 

<S S="Error">Add-DhcpServerv4Failover : Failed to verify if a failover relationship by the_x000D__x000A_</S>
   <S S="Error">name SFO-SIN-Failover exists on server 10.0.0.99._x000D__x000A_</S>
   <S S="Error">At line:1 char:74_x000D__x000A_</S>
   <S S="Error">+ if (-not(Get-DhcpServerv4Failover | ? { $_.Name -eq "MSDHCP0037Peer"})){ _x000D__x000A_</S>
   <S S="Error">Add-Dhc ..._x000D__x000A_</S>
   <S S="Error">+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~_x000D__x000A_</S>
   <S S="Error">~~~_x000D__x000A_</S>
   <S S="Error">+ CategoryInfo          : PermissionDenied: (MSDHCP0037Peer:root/Microsoft _x000D__x000A_</S>
   <S S="Error">/...erverv4Failover) [Add-DhcpServerv4Failover], CimException_x000D__x000A_</S>
   <S S="Error">+ FullyQualifiedErrorId : WIN32 5,Add-DhcpServerv4Failover_x000D__x000A_</S>
   <S S="Error">_x000D__x000A_</S>

the above powershell command  internally communicates to partner server and checks the same failover name if exists or not if exists it throws exception and if not it proceed to create, but the first step i.e connecting to partner server itself is not happening hence it is saying failed to verify causing the issue can we have any such refernces of adding the failover from other remote servers

Thanks





Clustering with Exchange 2007

$
0
0

Hi

We have an exchange 2007 server running on windows 2008 R2. A spare server recently became available and I have been asked to see if its possible to create a fail over cluster with our existing environment.

I appreciate that we have an un-supported version of exchange and windows but my boss wants to keep costs down so I want to want to see if its possible and if not what version do i need to go to. (Also he has ruled out virtualization)

I think its not possible as purchasing licences for old software could prove difficult.

Any feedback would be greatly appreciated.

Cheers

Colin

Failover cluster very slow

$
0
0

Hi everyone,

I took over admin tasks after one of admins left. He was responsible for failover cluster but I heard that users complained a lot. Main issue is that is very very slow. I logged in and check how it looks like.

It is 2 node cluster with witness disk.

Under network it has 2 networks both allowing cluster and client. There is no heartbeat network for cluster communiction. Both nics are in NIC Teaming mode. 

Storage - 2 csv are configured and each node is owner of 1.  Storage type is SAN and it is directly attached to the servers.

How to troubleshoot slow performance, what to think about or how to start troubleshooting?

IF you need more info just ask. Thank you all


File Cluster 2016 - Smb Share Permissions Update problem

$
0
0

Hi all,

I have file cluster setup on Windows Server 2016 environment with 4 nodes. Everything worked fine until now. I am not able to change share permission on one of the roles on all shared folders. The disk is connected via iSCSI and it is assigned to that role. I try with role re-creation, remove the disk and attach again, re-connect iSCSI connection for that disk. I found this topic, the guy who had the same issues with file cluster 2012 fix the issue with re-creation of file cluster, is that really necessary?

After few hours of troubleshooting i was able only to find this error in the event viewer:

Microsoft-Windows-FileServices-ServerManager-EventProvider/Operational 

Event ID: 1

Error: ERROR: Task 'example.ex.amp.le-Folder-SmbShareUpdate-4182a3a9-ecbe-4af6-96e7-12e7a1d1da36' has failed: Microsoft.FileServer.Management.Plugin.FSACException: Error occurred while updating an SMB share: The cluster resource could not be found.  ---> Microsoft.Management.Infrastructure.CimException: The cluster resource could not be found.
   at Microsoft.Management.Infrastructure.Internal.Operations.CimSyncEnumeratorBase`1.MoveNext()
   at System.Linq.Enumerable.SingleOrDefault[TSource](IEnumerable`1 source)
   at Microsoft.FileServer.Management.Plugin.Services.FSCimSession.ModifyInstance(ICimInstance instance)
   at Microsoft.FileServer.Management.Plugin.Providers.WmiUpdateTaskBase.DoWork(Object sender, DoWorkEventArgs e)
   --- End of inner exception stack trace ---
   at Microsoft.FileServer.Management.Plugin.Providers.WmiUpdateTaskBase.DoWork(Object sender, DoWorkEventArgs e)
   at System.ComponentModel.BackgroundWorker.OnDoWork(DoWorkEventArgs e)
   at System.ComponentModel.BackgroundWorker.WorkerThreadStart(Object argument)

Microsoft-Windows-FileServices-ServerManager-EventProvider/Operational

Event ID: 0

Exception: Caught exception Microsoft.Management.Infrastructure.CimException: The cluster resource could not be found.
   at Microsoft.Management.Infrastructure.Internal.Operations.CimSyncEnumeratorBase`1.MoveNext()
   at System.Linq.Enumerable.SingleOrDefault[TSource](IEnumerable`1 source)
   at Microsoft.FileServer.Management.Plugin.Services.FSCimSession.ModifyInstance(ICimInstance instance)
   at Microsoft.FileServer.Management.Plugin.Providers.WmiUpdateTaskBase.DoWork(Object sender, DoWorkEventArgs e)

This is the error in the file cluster management console:




Setting up Generic Services role on 2016

$
0
0
I'm a little confused on getting the generic services role setup on a failover cluster. I currently have a two node cluster configured. On each node, I have software that is not cluster aware but has directions on setting up HA with the generic services role. Where the confusion comes in is, why does the HA wizard require each service to have a name and IP address? I have a total of 6 services I need to run this way, does that mean I need to have IP's for each? Why does the generic service not run under the cluster name and IP? I've looked for better documentation on how to get this configured but there seems to be a lack good reference information on using the generic services role.  

Data Migration

$
0
0

Hi,

Server connected with local hard disks which has data about 5 TB. No SAN Attached. 

We need to move the data to different location. We cant ship the server. 

Dont want to do via ROBOCOPY. 

Please let me know other possibilities 

What is the work environment and culture like at Microsoft?

2 Node Cluster with local storage

$
0
0

I am not a server expert here, but I am a general admin, so I'll try to keep up ;) I know that versions of this question have been asked, but the answers vary. Hypothetically, I have a server, with a 1TB OS drive, and a 10TB data drive. We are going to run SQL and a proprietary application that keeps the Database updated. We want to make this a highly available server, so we have an identical server that we want to use (along with the first one) to create a two node cluster. 

From what I am reading, there are people saying that I cannot create a cluster using servers with local storage, and must use a SAN or NAS, others are saying yes I can. But there is no definitive answer I've found, and I do not want to spend the money to go from hypothetical to reality, if it can't be done.

The problem with a shared storage like NAS, is that it creates a single point of failure. If the NAS fails (power failure, board failure, fire, etc.), then the Database server cluster fails, and that defeats the whole point of the cluster. Clustering is all about eliminating single points of failure, so NAS is not the solution. SAN seems to me like clustered storage, and makes me wonder why I can't just eliminate the complication of the extra hardware, and utilize the two perfectly good arrays I already have in the existing nodes I'm clustering. A SAN in this case just seems like redundant overkill, and a waste of money. I already have the high availability of the two server cluster, and their identical internal storage arrays, with all the storage I need. Why can't I use them? 

Is there a way to cluster two servers, with their own locally attached storage? For example, can I take server A with a C: and a locally attached (eg internally configured RAID array) D: drive, and make server B (configured identically) work in a cluster with server A? And can it be made to work with the OS on C:, and the database on the internal D:, without a NAS or a SAN?

Go...


CLuster NIC Config for File Server

$
0
0

Building out a File server CLuster.

1.I currently have a NIC team to be used for file host

2. I have 2 sep NICs for ISCSI.

Should i create another team and use that as cluster only and potentially just IPV6 or is it best practice to just have the first NIC team as cluster and client and heartbeat will just go through that?

Drain role Failed

$
0
0

We have three Node N-1, N-2, N-3. I drain role from N-2 and there10 VM's Moved out of 14. 4 VM are not moving getting error . Tried to move manually but the error same. Please assists All the Node in are WIN-2012 R2

Error Message : "operation did not complete on resource virtual machine live migration"

How to assign cluster backup IP in Failover cluster.?

$
0
0
Assign cluster backup IP in Failover cluster, along with cluster service.

Cannot create checkpoint when shared vhdset (.vhds) is used by VM - 'not part of a checkpoint collection' error

$
0
0

We are trying to deploy 'guest cluster' scenario over HyperV with shared disks set over SOFS. By design .vhds format should fully support backup feature.

All machines (HyperV, guest, SOFS) are installed with Windows Server 2016 Datacenter. Two HyperV virtual machines are configured to use shared disk in .vhds format (located on SOFS cluster formed of two nodes). SOFS cluster has a share configured for applications and HyperV uses \\sofs_server\share_name\disk.vhds path to SOFS remote storage). Guest cluster is configured with 'File server' role and 'Failover clustering' feature to form a guest cluster. There are two disks configured on each of guest cluster nodes: 1 - private system disk in .vhdx format (OS) and 2 - shared .vhds disk on SOFS.

While trying to make a checkpoint for guest machine, I get following error:

Cannot take checkpoint for 'guest-cluster-node0' because one or more sharable VHDX are attached and this is not part of a checkpoint collection.

Production checkpoints are enabled for VM + 'Create standard checkpoint if it's not possible to create a production checkpoint' option is set. All integration services (including backup) are enabled for VM.

When I delete .vhds disk of shared drive from SCSI controller of VM, checkpoints are created normally (for private OS disk).

It is not clear what is 'checkpoint collection' and how to add shared .vhds disk to this collection. Please advise.

Thanks.

부천오피 ☞uuzoa2닷com☜ 〔유유닷컴〕 오피후기

CNO missing from ADUC and live migrations.

$
0
0
New to Failover clustering in Microsoft so this may be a fairly basic question. 

2 nodes in a failover cluster running Windows server 2016 and for some reason the CNO is not in ADUC (don't know whether it was ever there or whether it has been deleted as I did not build this cluster). Live migrations were failing so I built a test lab, replicated the exact problem and tried two different things - 

1) created an entry in ADUC in the computers group and then within FCM took the cluster server name offline, ran a repair and then brought it back online which allowed live migrations to work but in Server Manager -> All Servers it is still complaining about not being able to find the CNO (kerberos authentication error reported) and if I try to add it manually using "Add Servers" it cannot find the CNO in AD. 

I suspect this is because when the entry is created automatically when setting up a new failover cluster there is more going on than what I did by simply adding an entry manually. 

2) destroyed the cluster and recreated it which resulted in everything working correctly ie. not just live migrations but also the CNO is now displayed as online in Server Manager -> All Servers etc and the CNO entry was created automatically in ADUC computer group.

So my question is do I have to rebuild the cluster to avoid any future problems ie. if I do option 1) I can do live migrations but the fact that I am still getting errors in Server Manager and the CNO is not online makes me think I may face other issues further down the line. 

Is there a way to go with option 1 and resolve all the issues ? 

If not and rebuilding the cluster is the best thing to do are there any gotcha's I should look out for (my test lab rebuild went fine but the production setup is a lot more complicated). 

Many thanks for any pointers. 

Changing processor on existing Hyper V cluster 2012 R2

$
0
0

Dear All,

We have an 2 node Hyper V cluster, both nodes has 8 cores X 2 Processors, we have planned to upgrade the processor to 12 cores X 2 .

i just wanted to know will this have any issues with the existing cluster configuration, as i will be changing the processor on each node one by one.


TechGUy,System Administrator.


Building SOFS without DDNS

$
0
0

Please don't laugh, but yes, in the year 2019 I'm still running in an environment that does not have DDNS.

Does anyone know of a reference that shows what DNS records must exist to create a SOFS on Windows Server 2016?

TIA

Mike J ( stuck in the year 2000 )


MJ

S2D 2 node cluster

$
0
0

Hello,

We have 2 node S2D cluster with windows server 2019. Between two nodes we have directly connected RDMA storage network (Cluster only) and client-facing network based on LACP teaming on each node (Client And Cluster). We have done failover test and it works: when we power off one node, virtual machines migrates to another host as expected. But when we unplug client facing adapters (two adapters in LACP) on one node, where VM are resides, VM migration fails and after some time Cluster network name and Cluster IP address also failed. When we plug again client facing adapters (two adapters in LACP) to failed node, cluster IP address recover and VM client network works again. So the problem: cluster migration fails after unexpectedly shutdown of client facing network of one node, where VM are resides. Nodes can communicate with each other through Storage network and all nodes are up in Failover Cluster manager. So when client network is down, VM should migrate to another node with working client-facing network. But cluster fails and VM do not migrate. Where we can fix this behaviour? Has anyone met this before?

2-Node Stretch Cluster with Storage Replica?

$
0
0
I want to create a 2-node stretch cluster with Storage Replica.  I was able to setup the replica but once I create the cluster it breaks the storage replica.  I'm not sure if that is by design or not.  Is there something I am missing that prevents this type of scenario?  If I create the cluster first it does not allow me to create the replica with an error that it cannot find the volume on the source server that it must be a CSV on the cluster or added to a role on the cluster.  Neither of those are possible without the Replica running between the cluster nodes.

un-clustering hyperv nodes into standalone hosts

$
0
0

As the title suggests, I have a pair of server 2016 hosts which are currently operating as a 2-node failover cluster.

Due to relocation of resources, I'm looking to break up the nodes into two standalone servers. One of the servers is to be decommissioned from Hyper-V use, the other to remain as an operational stand-alone Hyper-V host.

Is there a known process for un-clustering nodes back to stand along hyper hosts.

I have a number of sizable VM's (storage size wise) which currently sit on the C:\ClusterStorage\ location, I'm trying to figure out if I need to create a new LUN on my SAN and migrate the VMs onto this new LUN, or whether I could utilize the existing disk LUN currently used by the CSV volume?

Is there scope for me to powerdown the VM's, uncluster the node, convert the CSV volume to a normal SAN based disk and power back up the VMs on the now standalone host?
*EDIT*
I know there's the option to 'remove from cluster shared volumes' but I'm not sure what the impact is of doing this is..






대구오피 “uuzoa2.com ” {유유닷컴} 오피가격

Viewing all 4519 articles
Browse latest View live


Latest Images

<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>