Quantcast
Channel: High Availability (Clustering) forum
Viewing all 3614 articles
Browse latest View live

SOFS and connections

$
0
0

Hey

I would like to create an active/active file server for User Profile Disks (vhdx)

I have create a SOFS cluster (Virtual machines + shared disk + CSV).

To my understandig the SOFS setup still uses one (disk owner) server to do the IO to the disk (shared disk on SAN)

If its true - how do I create an "real" active/active cluster? 

(Using Windows Server 2019)

I have approx 1000 concurrent connections....

Mike


Cluster startup

$
0
0

Suppose I have a 4 nodes cluster + 1 FSW. And at the beginning the 4 nodes are shutdown .

I would like to startup the first 2 nodes, however the cluster seems identify down. (maybe not enough quorum).

When I startup the 3rd nodes, the cluster is up.

May I know this is normal ? I suggest wonder the vote of FSW should also make the cluster have quorum .

 

Clustering between a physical server DASD and a VM

$
0
0

Scenario :

Physical Server (PS1) with DASD : WIndows 2012 R2 (Multiple terabytes of data on multiple partitions/drives)

Virtual Server 1 (VS1): Windows 2016 (target server Day 1)

Virtual Server 2 (VS2): Windows 2016 (target server Day 2, once all data is synced with VS1 PS1 will be decommissioned)

In the past I have created clusters but with new shared drives with 0 data on them on day 1. 

Now the challenge is to create a cluster without creating a new drive but to use an existing drive on PS1 and replicate the data to VS1 and then later to VS2 (in shared mode) and then decommission PS1.

All this without losing PS1 service availability and not erasing data at the source server.


Luis M Astudillo Freelance Enterprise/Infrastructure Architect and Technology Strategic Planner LinkedIn: www.linkedin.com/in/luisma

Simulate simultaneously failure nodes

$
0
0

Are there any way I can simulate simultaneously failure nodes? I use task scheduler to set shutdown at the same time, however

seems can't get my expected results. THe dynamic vote seems changing too fast/

Windows NLB takes too long to recover/failback

$
0
0

I've built a fresh Windows Server 2016 NLB cluster of two Web Application Proxy servers.

They are both Hyper-V VMs on the same host and with MACAddressSpoofing set to ON.

Both servers are dual-NIC and the cluster is using the second dedicated NIC in each case.

The Cluster is set to Unicast.

The cluster constantly responds to pings regardless of which node is down - this is good. 

BUT... the cluster fails to respond to HTTPS traffic for about a minute while either node is restarted, whether at the VM or NLB control level. As mentioned above, the ping is steady throughout.

Is this expected behaviour?

Storage Spaces Direct is the most unreliable, non highly-available high-availability solution Microsoft have ever released!

$
0
0

Storage Spaces Direct is the most unreliable, non highly-available high-availability solution Microsoft have ever released! We purchased a fully validated S2D solution from Dell and in two years since we put it in we have had:
> Total Cluster Failure with complete loss of data on CSV's requiring a full rebuild from scratch and complete DR of over 200 virtual machines!
> Hosts crashing during routine patching due to differences in the binaries causing CSVs to go offline and VMs to crash
> Bug after bug. 
> To fix the issues that were introduced in the May 2018 UR that in itself is a massive risk to install October UR that fixes it as during the patching cycle to install the Oct 2018 UR expect to see the loss of storage.
> There are still some serious bugs in S2D that Microsoft have been unable to fix and still no news.

> We have been advised by Dell that to patch S2D at present they for all customers they do not recommend carrying this out online and to either shutdown all VM roles on the cluster and patch or migrate all roles off the cluster to patch.

WTF!



Microsoft Partner

Windows 2016 DataCentre : Cannot Create Storage Replica for stretch cluster deployment.

$
0
0

I have 4 virtual machines running Windows 2016 with failover clustering.

These are spread across 2 data centers (2 VMs in each data center).

Each pair of VMs has 2 shared 'vhd sets' for Log and Replica Volumes

Cluster is configured and passes checks.

File server role is configured on site A and Proposed Replica Volume is selected.

When I try to replicate the volume with the file server I get the following error when the wizard tries to complete:

* Failed to create replication.

ERROR CODE: 0x80131500:

NATIVE ERROR CODE : 3 .

Invalid namespace

There is now firewall on these servers at the moment and there is also no firewall between the data centers.

wmi seems to be running on the nodes as well.

Cluster passes validation checks with some minor warnings about single network adaptor etc.


Clustering windows server 2016 Data Center

$
0
0

Hello,

I have been asked to Cluster 2 HP Proliant Servers 380 Gen 10.

On both servers Raid 1 (2 first Disk)  and Raid 5 (three last disk ) have been configured.

I have already installed windows server 2016 Data Center on both of them and Hyper V also installed.

The ILOs have been configured.

I need help to configure cluster with Server 1 and server 1 and connect them physically for full redundancy on these 2 Cisco 3850 Switches.

I need heart beat configured.

Can someone help me or direct me to a forum or blof where I can follow the step by step process to do the clustering and physically connect these devices.

Just a side note that later there will be a SAN added and it will be connected to both of those switches.

Thanks a lot

Abhi



Firewall block traffic of heart beat

$
0
0

I want to do some test in cluster to lost heartbeat, I want to use windows firewall to perform this .

what port I should block ?

I tried block UDP and TCP 3343, but seems still not working .

Guys caw we get the RegKey back for the passing WSSD on the windows server SSD cluster please ...

$
0
0

as per the official Microsoft position on windows server 2019 Data Centre 

"...When can I deploy Storage Spaces Direct in Windows Server 2019 into production?

Microsoft recommends deploying Storage Spaces Direct on hardware validated by the WSSD program. For Windows Server 2019, the first wave of WSSD offers will launch in February 2019, in about three months.

If you choose instead to build your own with components from the Windows Server 2019 catalog with the SDDC AQs, you may be able to assemble eligible parts sooner. In this case, you can absolutely deploy into production – you’ll just need to contact Microsoft Support for instructions to work around the advisory message. ..."


regards,

Alex 


How can we move the Quorum Disk from Node1 to Node2 ? - Windows 2012 R2 - Hyper-V Clustering

$
0
0

Hello,

We have created a cluster with 2 Nodes and created a role for File share. There are totally 3 Disks in the cluster, among three we have allocated 1 disk as Quorum Disk.

When Node1 is powered off, all the 3 disks are moving to Node2 automatically. But I would like to know how can we move the Quorum Disk from Node1 to Node 2 when both the nodes are active ?

We can move the 2 Disks from Node1 to Node2 while both the Nodes are Powered On, but through the same option I am unable to Move the Quorum Disk from Node1 to Node2 (Right Click on the Disk -> Move -> Select Node).

Kindly suggest on this  !!

Thanks & Regards,

Anoop Nair.


Anoop Nair

Windows 2012 R2 crashes during backup - Bug Check code 0x00000133

$
0
0
Hi,
We have an HP DL580G7 server, OS Windows Server 2012 R2 Datacenter Core with Hyper-V that is crashing repeatedly due to Bug Check code  0x00000133 . The error most occurs during backup. The server is one of the Hyper-V cluster nodes. Both servers have the same configuration (drivers, firmware, hotfix...), with the only difference that the problematic server is connected to the autoloader.
If you require more information please let me know.
Has anyone else had this issue, if so is there any resolve for it?

Thank

Understand Clustering

$
0
0

Hello Team,

Could you please help me with Clustering articles. I have seen other article of Ask DS , can you please suggest which one to start off so that I could work and troubleshoot the issue ?

Thanks..

2016 to 2008 R2 cluster

$
0
0

Can I add a Windows Server 2016 node to a Windows Server 2008 R2 cluster?


Reza Negarestani

Data move from ClusterA (iSCSI) to ClusterB (FC)

$
0
0

Have existing Server 2012R2 cluster with iSCSI storage, need to move to new Server 2019 cluster with FC storage.

Do I have it right that one could do any:

1) Downtime (VMs moved via LAN from old cluster storage to new)

2) Add new 2019 hosts to existing cluster & configure at least one 2019 host to have iSCSI & then do LM?

3) Add new 2019 hosts to existing cluster & configure at least one old 2012 R2 host to have FC & then do LM?

How well 2019 adds to 2012R2 cluster?

If 2 or 3 then after data move, remove 2012R2 hosts & upgrade cluster version?

Any other options?

Seb


Get-Volume returns all volumes within Windows Failover Cluster instead of just local

$
0
0
Hello all.

This is my first entry in the forums, so apologies if I miss something or have this in the wrong place.

I am using the Get-Volume command in PowerShell to return all the volumes located on the server I am running it from.

However, our servers are members of Windows Failover Clusters.

On one of our clusters it does what I would expect. We get a list of all the volumes on this particular node. On the other cluster we get a list of all volumes within the cluster.

Does anyone know of any setting in the windows failover cluster (or anywhere else) that could explain the difference in behavior.



In addition if I try to create a new volume using New-Volume (PowerShell) in the cluster that behaves as expected it works without issue.

If I try to create a New-Volume using New-Volume (PowerShell) in the cluster that shows all volumes I get the below error:



Failover clustering could not be enabled for this storage object.
Activity ID: {<blanked>}
    + CategoryInfo          : NotSpecified: (:) [New-Volume], CimException
    + FullyQualifiedErrorId : StorageWMI 46008,Microsoft.Management.Infrastructure.CimCmdlets.InvokeCimMethodCommand,New-Volume
    + PSComputerName        : <blanked>




Any help on this would be greatly appreciated.

Thank you.


error FailoverClustering-Manager - I don't have a cluster

$
0
0

Hi everybody.

a few days ago known a long series of errors as i said in the title ...
I checked that the associated service is disabled because i do not have a cluster but nothing changes.
I checked that there is no cluster configuration and they do not actually exist.
I still clean chache, old cluster configurations (although this machine has never seen a cluster)

As I have known of the strong slowdowns from a few days i noticed that it is concurrent with this error but I just can not get it back to normal.

The error is as follows:

-System
-Provider
[ Name]Microsoft-Windows-FailoverClustering-Manager
[ Guid]{11B3C6B7-E06F-4191-BBB9-7099FFF55614}
EventID4657
Version0
Level2
Task2
Opcode0
Keywords0x8000000000000000
-TimeCreated
[ SystemTime]2018-12-14T10:29:33.170788700Z
EventRecordID2448336
-Correlation
[ ActivityID]{5D23092D-9393-0005-2381-235D9393D401}
-Execution
[ ProcessID]2792
[ ThreadID]7864
ChannelMicrosoft-Windows-FailoverClustering-Manager/Admin
ComputerENPAPNAS03.Ente_Enpap.local
-Security
[ UserID]S-1-5-18
-EventData
Parameter1Get-ClusterNode
Parameter2Il Servizio cluster non è in esecuzione. Verificare che il servizio sia in esecuzione su tutti i nodi del cluster.

thanks for your supports


for the second time in 6 weeks a 6 Node S2D cluster goes down after a simple disk failure. How is that possible ?

$
0
0

We are using a 6 node DELL R740xd. The cluster passed the cluster validation test.

the node's did not reboot but the Cluster Shared Volume was off line and came only back to live after a 22! hour check...

after the CSV was back online we found all VM's were being moved onto one host. 

after examining the logfiles it looked like the cluster service removed the owner node off the CSV from the cluster instead of deactivating the disk, or the node containing the faulty disk.

any suggestions as to why this is possible. 

Peter

2 nodes fail over cluster and disk witness quorum fail on fail over test

$
0
0

Hi,

I have configured Windows Server 2012 R2 failover cluster with 2 nodes.

I set quorum mode to disk witness with mapped disk from SAN Storage which is connected to servers directly.

and configured heartbeat network with direct server connection.

there were no warnings and fails on cluster validation test so, I did failover test.

I pulled LAN cables from active node, the quorum disk tried to move to standby node but, it couldn't be online.

it seems to be 'Split brain' is it possible to solve this situation with disk witness quorum?

or it file share witness only the solution?

thanks,

When to run Cluster validation wizard on a new node

$
0
0
I have an existing production multi-site/subnet 2012 failover cluster and I'm going to be adding a new node per site. I'm just trying to verify when I run the Cluster Validation wizard, is it before I install the failover cluster feature on the new node or after the feature is installed?

Salvador Diaz III

Viewing all 3614 articles
Browse latest View live


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