KernSafe Technologies
 
Picture
Many iSCSI vendors struggle to educate and show customers that in real world scenarios, iSCSI is better for enterprise use. Luckily demand for iSCSI solutions is growing on daily basis now but it wasn't like that from the very beginning. When first iSCSI Initiators went out, they weren't a good example of use of iSCSI technology. They had many problems which was successfully preventing joy of simplicity that comes with it.

Fiber-Chanel

Nowadays fibre-channel is running at 8GBit while iSCSI at 10GBbit. When you will compare those two protocols now, you will see that there isn’t much difference between them. In the past, one of the advantages of fibre-channel was security since it required it to be on completely separated network and not be connected in any way to any other network. Now, iSCSI is running on same network and it is nothing unusual but if someone is concerned about security, he may consider buying separate switches that will physically separate it from normal network traffic.

From more technical point of view, fibre-channel operates as a layer-two protocol, while iSCSI operates on layer-seven (application) level. The biggest difference between that is that fibre-channel uses its own flow and error correction mechanism, while iSCSI depends on build in TCP/IP solutions.

Also processing of fibre-channel frame take a slightly less time since there doesn't need to be any TCP and IP header generated and decoded each frame is sent. Since the fibre-channel is roll out for a specific purpose, all frame management is performed in dedicated hardware and thanks to that it can operate with much less latency. At the same time iSCSI depends on TCP/IP in order to operate and encode and decode headers if required. It is still possible to offload some of this work (like generating checksums) into hardware that decreases CPU load but still some latency will be added.

Another thing is that fibre-channel is a frame lossless solution. If everything was correctly configured, there won't be any packet loss. Even if there will be huge increase of data transfer, flow control mechanism will limit amount of data being sent thus preventing any packet loss.
iSCSI

iSCSI is currently running at 10GBit speed but with 40GBit on mind, you can assume that it will be able to run any type of data, no matter how intense it would be.

We can also observe a slow adoption of 16GBit fibre-channels with 32GBit in future. But will it last? Companies that will decide to go for fibre-channel will need to be prepared that everything needs to be upgraded faster and faster - that comes with new expenses. At that time you may also consider FCoE (Fibre-channel over Ethernet). Then you can have approximately same speed but with hermetization of FC packages into Ethernet.

On the other hand Ethernet is not a lossless packet solution. It was design form the beginning as a 'best-effort packet delivery protocol' that means it was already known that some packets will get lost along the way. Said that, protocols (such as TCP) needs to control and detect if packets get lost and then be able to transfer lost packets again. This works in real life environment without any problems and, most importantly, there will never be any data corruption caused by any packet loss. Keep in mind if there will a large quantity of data being transfer over such network, its performance may drop significantly.

iSCSI may seems like it have some disadvantages but it can easily make them up with its simplicity and flexibility. On the contrary to fibre-channel, iSCSI don’t need any special network to work, it can utilize any Ethernet network. Actually its traffic can be routed outside of a single domain. That thing wouldn’t be possible with fibre-channel without any expensive hardware.

Simplicity is also very big plus for iSCSI. Setting up fibre-channel usually requires a special knowledge, while doing same task with iSCSI requires little to none additional knowledge to that network administrator already have. Even additional tweaking of it, in order to get its full potential would be far easier than in fibre-channel.

Fibre-channel over Ethernet (FCoE)

It is the newest technology that allows using fibre-channel lossless solutions over Ethernet network. It will need to use some special extensions like DCB (Data Center Bridging) to work. That will almost always require special hardware and configuration to work but it allows combining fibre-channel and Ethernet networks. Even if you probably can't use FCoE over your current network and you will need to cover some expenses to implement it, it still brings it closer to iSCSI then to fibre-channel in terms of price. The only disadvantage of it is that it is still a layer-two protocol and that means its traffic can't be routed outside broadcast domain. That makes managing of FCoE closer to fibre-channel rather than iSCSI.

What to expect

What can future bring that is still not here yet? I will still stick to iSCSI since its future is clearer comparing to fibre-channel. The speed of network will significantly increase to 100GBit in upcoming years. Fibre-channel future maybe is not that exciting but it is still something to look for.

Conclusion

Even if iSCSI will have some disadvantages, such as higher latency and packet loss, overall it is clear that every company can benefit from implementing it instead of fibre-channel because of its ease of deployment and flexibility of use. Already many companies decided to go for iSCSI and they are happy about it, not only for themselves, but also for their customers.

 
Picture
iStorage Server supports creation of Failover Clustering

Cluster Shared Volumes (CSV) is a feature of Failover Clustering available in Windows Server 2008 R2 for use with the Hyper-V role. A Cluster Shared Volume is a standard Cluster disk containing an NTFS volume that is made accessible for read and write operations by all nodes within the cluster. This gives the virtual machine (VM) complete mobility throughout the cluster as any node can be an owner, and changing owners is easy. In Windows Server 2008, the VM and all dependent LUN resources form a dependent group that can only be moved or failed over as a complete unit. This means that if additional VMs created dependencies on these same LUNs they become members of that same group and cannot be independently moved. Move one VM or LUN and you end up moving all of the VMs and LUNs in the group. Additionally, this model created storage management challenges where hundreds of VMs would require hundreds of LUNs. This all changes in Windows 2008 R2 with CSV. Now, the VMs with VHDs deployed on a single CSV have no dependency on the disk resource and can be freely moved to or from any node in the cluster.

To understand how Cluster Shared Volumes (CSV) works in a failover cluster, it is helpful to review how a cluster works without CSV. Without CSV, a failover cluster allows a given disk (LUN) to be accessed by only one node at a time. Given this constraint, each Hyper-V virtual machine in the failover cluster requires its own set of LUNs in order to be migrated or fail over independently of other virtual machines. In this type of deployment, the number of LUNs must increase with the addition of each virtual machine, which makes management of LUNs and clustered virtual machines more complex.

Cluster Shared Volumes provides the following benefits in a failover cluster:
  • The configuration of clustered virtual machines is much simpler than before.
  •  You can reduce the number of LUNs (disks) required for your virtual machines, instead of having to manage one LUN per virtual machine, which was previously the recommended configuration (because the LUN was the unit of failover). Many virtual machines can use a single LUN and can fail over without causing the other virtual machines on the same LUN to also fail over.
  • You can make better use of disk space, because you do not need to place each Virtual Hard Disk (VHD) file on a separate disk with extra free space set aside just for that VHD file. Instead, the free space on a Cluster Shared Volume can be used by any VHD file on that volume.
  •  You can more easily track the paths to VHD files and other files used by virtual machines. You can specify the path names, instead of identifying disks by drive letters (limited to the number of letters in the alphabet) or identifiers called GUIDs (which are hard to use and remember). With Cluster Shared Volumes, the path appears to be on the system drive of the node, under the \ClusterStorage folder. However, this path is the same when viewed from any node in the cluster.
  • If you use a few Cluster Shared Volumes to create a configuration that supports many clustered virtual machines, you can perform validation more quickly than you could with a configuration that uses many LUNs to support many clustered virtual machines. With fewer LUNs, validation runs more quickly. (You perform validation by running the Validate a Configuration Wizard in the snap-in for failover clusters.)
  • There are no special hardware requirements beyond what is already required for storage in a failover cluster (although Cluster Shared Volumes require NTFS).
  • Resiliency is increased, because the cluster can respond correctly even if connectivity between one node and the SAN is interrupted, or part of a network is down. The cluster will re-route the Cluster Shared Volumes communication through an intact part of the SAN or network.

On more information on how to deploy CSV in Windows Server 2008 R2, please refer to this site http://blogs.msdn.com/b/clustering/archive/2009/02/19/9433146.aspx


Picture
Picture
Picture
You can download Trial version of iStorage Server here.
 
Picture
iStorage Server supports High-Availability iSCSI SAN for Windows Server 2008 & Hyper-V Clustering

iSCSI,  an Internet Protocol (IP)-based storage networking standard for linking data storage facilities. By carrying SCSI commands over IP networks, iSCSI is used to facilitate data transfers over intranets and to manage storage over long distances. iSCSI can be used to transmit data over local area networks (LANs), wide area networks (WANs), or the Internet and can enable location-independent data storage and retrieval.  It is a popular Storage Area Network (SAN) protocol, allowing organizations to consolidate storage into data center storage arrays while providing hosts (such as database and web servers) with the illusion of locally-attached disks. Unlike traditional Fibre Channel, which requires special-purpose cabling, iSCSI can be run over long distances using existing network infrastructure.

High-availability clusters (also known as HA Clusters or Failover Clusters) are computer clusters that are implemented primarily for the purpose of providing high availability of services which the cluster provides. They operate by having redundant computers or nodes which are then used to provide service when system components fail. Normally, if a server with a particular application crashes, the application will be unavailable until someone fixes the crashed server. HA clustering remedies this situation by detecting hardware/software faults, and immediately restarting the application on another system without requiring administrative intervention, a process known as Failover. As part of this process, clustering software may configure the node before starting the application on it. For example, appropriate file systems may need to be imported and mounted, network hardware may have to be configured, and some supporting applications may need to be running as well.

The Hyper-V role enables you to create a virtualized server computing environment using a technology that is part of the Windows Server 2008 operating system. This solution is provided through Hyper-V. You can use a virtualized computing environment to improve the efficiency of your computing resources and improve server availability without using as many physical computers as you would need in a failover configuration that uses only physical computers.

Main benefits of Hyper-V are:
  • Consolidation of hardware resources
  • Ease of administration
  • Significant cost savings
  • Fault tolerance support through Hyper-V clustering
  • Ease of deployment and management
If you want to learn more about iStorage Server support for High-Availability iSCSI SAN and Hyper-V under Windows Server 2008 operating system, please refer to this White Paper.

Picture
Picture
Picture
You can download Trial version of iStorage Server here.