Search This Blog

Friday, October 8, 2010

Different Generations of Dell Poweredge 1950 Server

The Dell Poweredge 1950, although one generation removed from the Dell line of Poweredge Servers, is still in high demand on the used market.

The 3 Generations of Poweredge 1950 series allows you to perform a myriad of duties in any modern data center. From application servers, to populating databases, from file servers to print servers, from mail services to collaboration services, the modern server is indeed to backbone of any modern business, be it large or small.

Below are the basics of each of the PE1950 server complete with generation. This can assist if your software calls for specific requirements. Again this doesn’t drill down to every bit of difference between the generations, but a guideline.

Dell PE1950 1U Gen I

Processors: Up to 2, Intel Dual Core processors; Max processor speed is 3.73, max cache is 4m, max front side bus is 1333Mhz.

Memory: Up to 32gb 667Mhz

I/O Slots: three expansion slots; 2 x 64-bit PCI-X; 1 x PCI Express x8.

Drive Controllers: Embedded PERC integrated PCI SAS controller.

Raid Controllers: Embedded PERC 5/i, PERC4e/Di (single channel RAID with 256MB of battery-backup cache)

Drive Bays: 4, 2.5” or 2, 3.5” SAS or SATA Hard drives

Power supplies: 750W hot-plug/redundant.

Dell PE1950 1U Gen II

Processors: Up to 2, Intel Dual Core processors; Max processor speed is 3.0, max cache is 8m, max front side bus is 1333Mhz.

Memory: Up to 32gb 667Mhz

I/O Slots: three expansion slots; 2 x 64-bit PCI-X; 1 x PCI Express x8.

Drive Controllers: Embedded PERC integrated PCI SAS controller.

Raid Controllers: Embedded PERC 5/i, PERC4e/Di (single channel RAID with 256MB of battery-backup cache)

Drive Bays: 4, 2.5” or 2, 3.5” SAS or SATA Hard drives

Power supplies: 750W hot-plug/redundant.

Dell PE1950 1U Gen III

Processors: Up to 2, Intel Dual Core processors; Max processor speed is 3.0, max cache is 12m, max front side bus is 1333Mhz.

Memory: Up to 64gb 667Mhz

I/O Slots: three expansion slots; 2 x 64-bit PCI-X; 1 x PCI Express x8.

Drive Controllers: Embedded PERC integrated PCI SAS controller.

Raid Controllers: Embedded PERC 5/i, PERC4e/Di (single channel RAID with 256MB of battery-backup cache)

Drive Bays: 4, 2.5” or 2, 3.5” SAS or SATA Hard drives

Power supplies: 750W hot-plug/redundant.

About velocityanne
Charming and personable sales GURU for Velocity Tech Solutions. We are the BEST source for Dell Poweredge and HP Proliant servers and server parts.
View all posts by velocityanne →

Thursday, October 7, 2010

Types of Raid and the Advantages of Each

I thought I’d let our head Geek write more words of her server wisdom. We get so many questions around setting up raid on Dell Poweredge Servers that we thought this would be an easy way to help.

What are the types of Raid and the advantages of each?

By Kay Winchell, Velocity Tech Solutions, www.velocitytechsolutions.com

A vitally important function of the network administrator is to protect and secure the data of a business. An equally important function is maximizing the performance of the network. A cost effective way of both protecting data and maintaining access to the network is to install a RAID array on your server’s hard drives. RAID, or Redundant Array of Independent Disks, is what we use to describe a data storage method that can manage data among multiple hard disk drives on a server . The main reason a Raid Array is used is for performance and redundancy, the benefits of which depend on the type of Raid chosen.

The most common types of Raid Arrays for servers are Raid 0, Raid 1, and Raid 5, 6, 10, 50 and 60. We will discuss the trade-offs between performance (speed of access) and redundancy (protection of data) with each type of RAID.

Raid 0, which is also called Striping, provides for highest performance and zero redundancy. The Raid controller writes data across all drives in the array and reads and writes to multiple disks at the same time. Speed is enhanced, however, if one hard drive in the array fails, total data loss occurs. A business will usually use Raid 0 when speedy temporary access to large capacity disk space is needed, and in the case of disk failure the data can be easily reloaded from another source without impacting the business. 100% of the disk space is available for use with a RAID 0 configuration, so the cost of storage is the lowest. To summarize this option, speed is high, disk storage cost is the lowest, and safety of the data is lowest.

Raid 1, which is also called Mirroring, provides for high redundancy, but zero increase in performance. Each hard drive is paired with another, one being a complete copy, or mirror, of the other. If one hard drive fails, the paired drive contains the same data. To set up Raid 1, the server must have at least two hard drives. Businesses will usually use a RAID 1 when absolute data safety and access are required. However, cost of disk storage is a secondary consideration, as only 50% of the disk space is actually available. To summarize this option, speed is medium, disk storage cost is the highest, and data safety is high.

Raid 5, which is also called Stripe Sets with Parity, has a mix of performance, storage cost, and redundancy benefits. Data is written to all of the drives in the array and parity data is written to all of the drives in the array as well. The result does not require disk duplication like mirroring but does maintain redundancy. If a hard drive in the array fails, a new drive can be added and the array repairs itself while the system continues to operate normally. For this reason, Raid 5 is the array of choice for most servers with three or more disks. The RAID parity requires one drive per RAID set, so disk availability is always one hard drive less than the number of drives in the array, still much better than the 50% capacity of mirroring.

Raid 6, which is also called Stripe Sets with Dual Parity, is designed to improve the disk failure tolerance of RAID 5 by withstanding the failure of two drives in the array. Raid 6 functions like Raid 5, with stripe sets and parity, but adds a second parity scheme that distributes data across different disks than the first parity scheme to enable dual parity. Raid 6 requires a minimum of 4 disks and the parity requires two hard drives per RAID set. Raid 6 is a particularly good option for servers that use SATA drives, which are less expensive, but also less reliable, and may possibly require the ability to withstand a double disk failure.

Raid 10, which is also called Stripe Sets and Mirroring, is a multiple array set that combines Raid 0 Striping and Raid 1 Mirroring. The base level is two or more Raid 1 arrays, which as you will recall, is an array where each drive has an exact duplicate of itself. The second level is a Raid 0 array which stripes all of the data from the sub-arrays across all of the drives in its array. So, there is mirror redundancy in the sub-array and no redundancy in the main array. This provides for speed of access and combined capacity in the main array. One drive can fail in each sub-array and still be recoverable. Like mirroring, only 50% of the total drive capacity is available, but both redundancy and performance are enhanced by this Raid method.

Raid 50, which is also called Stripe Sets, is another multiple array set that combines Raid 5 with Raid 0. The base level is two or more Raid 5 arrays, which as you will recall, uses stripes and parity to maintain redundancy. The second level is a Raid 0 array which stripes all of the data from the sub-arrays across all of the drives in its array. So, there is parity redundancy in the sub-array and no redundancy in the main array. This provides for speed of access and combined capacity in the main array. One drive can fail in each sub-array and still be recoverable. However, like the difference between Raid 1 and Raid 5, the total disk capacity is increased because parity uses only one drive per array, as opposed to mirroring, which uses one drive out of each two.

Raid 60, which is also called Stripe Sets and Dual Parity Stripe Sets, is another multiple array set that combines Raid 6 with Raid 0. The base level is two or more Raid 6 arrays, which as you will recall, uses dual parity stripe sets to maintain redundancy. The second level is again a Raid 0 array, which stripes all of the data from the sub-arrays across all of the drives in its array. So, there is dual parity redundancy in the sub-array, and no redundancy in the main array. Again, this provides for speed of access and combined capacity in the main array. Two drives can fail in each sub-array and still be recoverable. Total disk capacity is still increased over mirroring, while preserving redundancy.

Your server, depending on its age, may not be capable of installing the newer types of Raid array.