Even after 72 hours of reading MANY articles, guides, and tutorials, along with watching countless YouTube videos and Google-ing like a madman, I still apparently understand nothing about failover clusters since I cannot get this to work for the life of me. So as you can tell from my demeanor, I am becoming VERY agitated to the point that I am ready to pull all my hair out. So now that I have gotten my rant out of the way, here is my situation:
I am trying to setup SQL Server 2012 in a dual-server environment for Team Foundation Server 2012. All of my servers are currently VMs in Hyper-V, which will eventually be migrated either to physical machines or other physical Hyper-V servers depending on the resources needed to sustain them.
The entire design of my infrastructure is to sustain the failure of any single physical machine. So for example, I have two domain controllers: DC-01 and DC-02. These two VMs are currently both running in my testing environment on the same physical machine, but they will eventually be placed on separate physical Hyper-V servers. That way, if one of those physical machines went down for any reason, the other domain controller would still be available since it was placed on a separate Hyper-V server.
So the model for my database servers is exactly the same. I have two database servers: DB-01 and DB-02. They are also both currently running in my testing environment, but will eventually be placed on separate physical Hyper-V servers as well. Now, the title of my post is actually a little misleading, because when it comes to clusters themselves, I have no problem setting them up. It's really not that difficult to do. I just install the Failover Clustering feature on both nodes, DB-01 and DB-02. Then after I validate and create the cluster, which I subsequently name DB-C1, all is well, right? No!
While the cluster validation technically passed, it never finds any valid cluster storage, but these only produce warnings. So I can create a cluster just fine, but these clusters don't have any storage. So now when I try to run the SQL Server 2012 setup for a SQL Server Failover Cluster, it fails on the "Cluster shared disk available check".
So while there is plenty of documentation out there on how to create a failover cluster, none of these stupid articles actually provide any explanation about the cluster storage. At first, you could tell I knew absolutely nothing about cluster storage since I tried a bunch of random tasks like adding secondary disks to DB-01 and DB-02, or tried to create storage shares or storage pools. But nothing actually worked as an available storage disk for my database cluster.
So then I assume the problem is that I need a file server cluster, so I start wasting my time creating two more servers: FS-01 and FS-02. I put them in a file server cluster called FS-C1, hoping that this would somehow remedy the situation, but I can't even figure out how to setup a clustered file server because I run into the same problem, and after further research, it appears that even if I had a clustered file server, it wouldn't even help to solve my database problem anyway.
From what I understand, my DB-C1 cluster needs a storage area network (SAN), iSCSI, SMB 3.0, or some other equally stupid technologies, all of which I know absolutely nothing about.
So after this extremely long explanation of my problem (by the way, thank you for reading everything so far), here is my question:
How can I setup FS-01 and FS-02 to work as a shared storage for my DB-C1 cluster?
From the information I have gathered so far (which I have most likely misinterpreted), the shared storage will be a single machine with some disks or something like that. However, this is not what I want, because it doesn't follow my infrastructure model. If I setup a shared storage on some single machine, if that machine went down, then my database servers would be absolutely useless anyway, right? And if so, how can I set it up where this clustered storage is actually itself a failover shared storage, so that if one of the file server VMs that the shared storage resides on were to fail, it could failover to the other file server VM on a separate physical machine? Because in the end, that's all I really want here; to just setup all these systems in such a way that if one of my Hyper-V servers lost power, crashed, or lost network connectivity, whatever the problem may be, it will not bring down my infrastructure.
And finally, keep in mind, that since I understand absolutely nothing about this shared storage, I most likely don't have any of the hardware necessary to set it up based on what I have read. So whatever the solution may be, it will have to be through Windows Server 2012, since that's all I really have. So I assume I'm going to be setting up an iSCSI Target Server or something along those lines, but again, I apparently have no idea what I am doing here, so that is most likely wrong as well.
Anyway, if you ended up reading all of this and can provide any insight to help me setup this infrastructure, I would GREATLY appreciate it, because I am totally losing my mind here, and because of all these headaches, I have gotten very little sleep, and simply want to get this done already!
Regards,
Nick
I am trying to setup SQL Server 2012 in a dual-server environment for Team Foundation Server 2012. All of my servers are currently VMs in Hyper-V, which will eventually be migrated either to physical machines or other physical Hyper-V servers depending on the resources needed to sustain them.
The entire design of my infrastructure is to sustain the failure of any single physical machine. So for example, I have two domain controllers: DC-01 and DC-02. These two VMs are currently both running in my testing environment on the same physical machine, but they will eventually be placed on separate physical Hyper-V servers. That way, if one of those physical machines went down for any reason, the other domain controller would still be available since it was placed on a separate Hyper-V server.
So the model for my database servers is exactly the same. I have two database servers: DB-01 and DB-02. They are also both currently running in my testing environment, but will eventually be placed on separate physical Hyper-V servers as well. Now, the title of my post is actually a little misleading, because when it comes to clusters themselves, I have no problem setting them up. It's really not that difficult to do. I just install the Failover Clustering feature on both nodes, DB-01 and DB-02. Then after I validate and create the cluster, which I subsequently name DB-C1, all is well, right? No!
While the cluster validation technically passed, it never finds any valid cluster storage, but these only produce warnings. So I can create a cluster just fine, but these clusters don't have any storage. So now when I try to run the SQL Server 2012 setup for a SQL Server Failover Cluster, it fails on the "Cluster shared disk available check".
So while there is plenty of documentation out there on how to create a failover cluster, none of these stupid articles actually provide any explanation about the cluster storage. At first, you could tell I knew absolutely nothing about cluster storage since I tried a bunch of random tasks like adding secondary disks to DB-01 and DB-02, or tried to create storage shares or storage pools. But nothing actually worked as an available storage disk for my database cluster.
So then I assume the problem is that I need a file server cluster, so I start wasting my time creating two more servers: FS-01 and FS-02. I put them in a file server cluster called FS-C1, hoping that this would somehow remedy the situation, but I can't even figure out how to setup a clustered file server because I run into the same problem, and after further research, it appears that even if I had a clustered file server, it wouldn't even help to solve my database problem anyway.
From what I understand, my DB-C1 cluster needs a storage area network (SAN), iSCSI, SMB 3.0, or some other equally stupid technologies, all of which I know absolutely nothing about.
So after this extremely long explanation of my problem (by the way, thank you for reading everything so far), here is my question:
How can I setup FS-01 and FS-02 to work as a shared storage for my DB-C1 cluster?
From the information I have gathered so far (which I have most likely misinterpreted), the shared storage will be a single machine with some disks or something like that. However, this is not what I want, because it doesn't follow my infrastructure model. If I setup a shared storage on some single machine, if that machine went down, then my database servers would be absolutely useless anyway, right? And if so, how can I set it up where this clustered storage is actually itself a failover shared storage, so that if one of the file server VMs that the shared storage resides on were to fail, it could failover to the other file server VM on a separate physical machine? Because in the end, that's all I really want here; to just setup all these systems in such a way that if one of my Hyper-V servers lost power, crashed, or lost network connectivity, whatever the problem may be, it will not bring down my infrastructure.
And finally, keep in mind, that since I understand absolutely nothing about this shared storage, I most likely don't have any of the hardware necessary to set it up based on what I have read. So whatever the solution may be, it will have to be through Windows Server 2012, since that's all I really have. So I assume I'm going to be setting up an iSCSI Target Server or something along those lines, but again, I apparently have no idea what I am doing here, so that is most likely wrong as well.
Anyway, if you ended up reading all of this and can provide any insight to help me setup this infrastructure, I would GREATLY appreciate it, because I am totally losing my mind here, and because of all these headaches, I have gotten very little sleep, and simply want to get this done already!
Regards,
Nick
Last edited:
Facebook
Twitter