iSCSI vs Fibre Channel.. Battle Continues..

December 28, 2011

There was a time that some claimed that iSCSI is not mature and is not suitable to serve the storage needs of high demanding data centers.

Those days have passed, and even those companies who promoted Fibre Channel, are starting to offer iSCSI products. There is no argument anymore that iSCSI has taken hold as the interface of the choice for new storage installations. All recent market studies have shown that Fibre Channel storage market has been either steady or has had a slight year to year improvement, while iSCSI storage market has been on the double digit rise year after year.

All features that have been once only available on high end fibre channel SANs are now available in iSCSI storage products. All the arguments about performance advantages of Fibre Channel are disappearing since 10G iSCSI has started shipping in numbers. And most important of all, there is no argument anymore that iSCSI storage costs are a lot less thanaFibre Channel SAN with equivalent performance and features.Only those companies who do not have iSCSI storage are still trying to promote Fibre Channel. But users are smart and do not see any compelling argument anymore to buy Fibre Channel SAN.

Having said all that, we do not think Fibre Channel is going to die anytime soon. As like any other technology, it takes a long time for an established technology to wind down and get replaced by another one. SCSI storage is still available when some predicted it will die 10 years ago. But from all market indications it is obvious that storage trends are moving toward IP Storage rather than Fibre Channel.

In summary, iSCSI, when properly implemented, provides high performance, hence making it a suitable to deploy in core applications. iSCSI is easy to understand and lower cost makes it a strong choice for medium to large enterprises. Having an end-to-end Ethernet IP environment is an added benefit to iSCSI. A properly designed iSCSI infrastructure and storage platform is good enough to handle any storage application currently handled by a Fibre Channel SAN.


Storage Tiering- What is it and what does it give me?

December 28, 2011

What is it?

Since there are many disk drives types, it is very important to optimize use of different drive types to achieve best price/performance ratio for your storage. Storage Tiering allows usage of different type of drives for different applications.

Common disk drives used in storage subsystems such as Stonefly’s IP Storage products are: SATA HDD, SAS HDD, and SSD. SATA and SAS HDDs are both mechanical disk drives with spinning platters. SSD stands for Solid State Drive and is made from non-volatile flash memory. HDDs come in different rotation speeds ranging from 5400 RPM (Rotation Per Minute) up to 15000 RPM.

 

In storage Tiering, the most reliable high- performance equipment is used to support the most critical data. Whereas, the most cost-effective resources to support older, less critical information

 

All three drive types are available with:

–      3 or 6Giga-Bit-Per-Second per port

–      Single port which is commonly called SATA interface, or SAS interface which is commonly called SAS interface.

All three drive types come in different capacities of up to 3Tera-Bytes, and these capacities keep increasing each year.

 

This many variations in drive type, capacity, interface, and speed results in large differences in price for each drive.

 

Storage companies offer wide variety of drive types to accommodate needs of different customers with different applications. Storage Tiering allows customer to have multiple drive types in their storage units, and gives them the ability to allocate certain type drives to certain applications depending on what that application needs.

What does it give me?

Answer is simple: Freedom to choose drive types depending on application, and overall reduction in data storage costs.


Title: Thin provisioning. What is it and is it worth paying money for it?

December 23, 2011

What is it?

 

Thin Provisioning is a technique used to optimize and efficiently utilize the available space in SAN. Stonefly’s IPSANs currently feature this technology.

Thin provisioning is mainly to eliminate over allocation of storage and increase storage utilization.

 

Conventionally, storage provisioning, also known as fat provisioning (FP), was done, where storage space was allocated beyond how much was needed. This resulted in low utilization rates and led to a lot of wastage of storage space. Users did not know how much storage they need, so they tend to ask for a very large # just to make sure they never run out of space. That resulted in over allocation of capacity.

 

Thin provisioning dynamically allocates storage to the clients from one unused storage pool. Users can ask for any capacity they want, but they only will use the capacity that they need. And as their need grows, so does their storage allocation.

 

In advanced thin provisioning implementations such as stonefly’s thin provisioning, there is also a feature called “Space Reclamation”. This is used if a large thin storage get allocated to a client, and client then deletes some of that data users can reclaim that deleted space to bring that back to the free storage pool.

 

In summary, Thin Provisioning provides the following advantages:

–      Allows you to add storage when necessary without downtime

–      Eliminates problems of over-allocation, hence providing optimal storage space utilization.

–      Reclaim your capacity after data is deleted

–      Reduced consumption of electricity, and hence reduced heat generation.

 

Is it worth paying money for it?

 

The costs involved in implementing and managing Thin provisioning are well worth it. It reduces storage costs and accelerates the returns on investment.

Studies have shown in large data centers with traditional storage, there is always about 30 to 50% unused capacity allocated to the clients. That space is paid for, but never used. Using thin provisioning you can reduce your storage need by about 30 to 50%. So using thin provisioning, you can reduce your storage costs by about 30 to 50%. Considering this saving, answer to this question is very obvious.


Why Shared folder encryption makes sense, despite Full disk encryption.

December 21, 2011

Firstly, Shared folder encryption means when files and folders are shared over networks, they are protected using encryption rules and protocols, hence securing sensitive data being shared.

Full Disk encryption is when the entire hard drive is encrypted sector- by – sector, rather than just encryption certain files or folders. The contents of the physical hard disk are encrypted. In Full disk encryption, separate protection rules are not defined for specific files, reducing the risk of leaving temporary files, et al unprotected.

StoneFly supports Block level encryption:  This type of encryption is applied to all data files, so that no data files are accidentally skipped. This is one of the most preferred option for data encryption as there is less room for mistakes with this type.

When it comes to encryption of data, there are pros and cons, Full disk encryption is good in a case when your laptop is stolen, but when u consider sending sensitive information over a network, you would need to rely on a different protection scheme. Sometimes, different encryption products are needed to be used at the same time for full protection of data.


How do you measure and report Service uptime?

December 21, 2011

Firstly, Uptime can be defined as the time a service/machine has been efficiently rendering service/reliable and stable, without any interruptions or break in service.

Uptime of a service also depends on the availability of that service, because, only when the service is available, it can be used efficiently.

Uptime can be calculated as a difference between availability and downtime of a service.

Uptime = Availability- Downtime

Where

“Availability” = A measure in time where a service is capable of being used.

“Downtime” = When the service is available but gives unreliable service or is not optimal/ has interruptions.

The uptime of a service can also be measured based on customer response and ratings.

StoneFly’s resources/casestudies  provides interesting information about service uptime.

Follow the link provided below:

http://www.stonefly.com/resources/casestudies_detail.asp?id=121


Will Cloud Storage Replace Local Storage?

December 20, 2011

With Cloud Storage gaining hype in providing storage solutions and recent developments in its technology, there is a mis-conception that local storage will soon become outdated.

As much as Cloud storage is useful for some applications, local storage is preferred for applications that demand better storage options. This is so because large enterprise applications need high speed access to their storage, which is not practically possible by cloud storage due to the low WAN access speed to the cloud.

Also, in Local storage, you are in control of your data, it is accessible without an internet connection and more secure when compared to the data stored on a cloud.

Although cloud storage is perfect for enterprise computing, local storage benefits outweigh cloud storage.

For more information on the local and cloud storage solutions, visit StoneFly’s iSCSI.com.

The link below talks about StoneFly’s product, Windows ESS, giving you a better understanding on how Local as well as Cloud storage solutions can be uses as per the enterprise’s requirement.

http://www.iscsi.com/products/software/ess/


What is snapshot, why do I need it?

December 19, 2011

Currently, most storage providers use this technology called Snapshot. This technology uses a method called “Copy On Write” which means that, if a location is changed, old data from the time snapshot for that location is stored in a repository called SnapSpace. The consequential snapshot will look like an actual volume, containing data at the time of snapshot taken.

Snapshot is also called Shadow Copy in windows Terminology. But the difference between windows based shadow copy and storage based snapshots are that windows based shadow copy will take the CPU cycles of the server, but storage snapshot does not.

Why do I need it?

Below listed are some usages of Snapshot.

–       Backup: Take a snapshot, backup the snapshot.

–      File Recovery: In case a file is deleted accidentally, it can be retrieved by going back to the snapshot.

–       Take a snapshot, give that to one or more other groups to work on it, while live volume is changing.

For more details on StoneFly’s Snapshot, Visit StoneFly’s  iscsi.com

Also refer the link given below:

http://www.iscsi.com/news/pressrelease_detail.asp?id=265