What is a RAID Controller?

RAID stands for Redundant Array of Inexpensive Disks. It is a technique used to store data on two or more disks in order to ensure data redundancy (in case of physical disk failures) and/ or improved storage read-write performance (as the host system can access data stored in multiple disks simultaneously, thereby speeding up the process).

All the physical disks that combined to form a RAID configuration appears as a single disk to the host and is referred to as a RAID Array. One can create, configure and manage RAID Arrays using a RAID Controller.

There are two types of RAID Controllers – Hardware based RAID Controllers and Software based RAID Controllers. Hardware based RAID Controllers are recommended because they come with their own processing/ memory capacity which they use for RAID operations. But Software based RAID Controllers use the processing/ memory capacity of the host system (PC/ Server) they are based on, which reduces the system resources that can otherwise be used by applications.

Hardware RAID Controllers can be added to the PC/ Server (using a PCI-Express slot, for example) or they can come embedded on the system’s motherboard.

Salient points you need to know about RAID Controllers **

1. It is recommended to have similar capacity hard disks within the same RAID array. Multiple sized hard disks can be used within the same array, but the maximum usable capacity per disk would be restricted to the smallest disk on the array.

2. A few RAID Controllers can support both internal and external storage disks.

3. If a system supports hot-swapping of disks, it is possible to replace a failed disk in the RAID array without switching the system off. After the new disk is inserted, the RAID controller begins to rebuild the RAID array using the new drive (automatically, in most cases).

4. It may take up to 15 minutes per GB during the rebuilding process. But this time can vary based on a lot of other factors as well.

5. A spare disk drive can be installed in a hot-standby mode within the host system so that upon failure of one of the disks in a RAID Array, it can automatically (and immediately) be used to reconstruct the data in the failed disk and hence the RAID array is reconstructed without data loss.

6. Even if the RAID Controller fails, it is possible to replace it with a new one without losing data on the disk drives as long as both of them save configuration data in the same way.

7. Some RAID Controllers may have in-built batteries to preserve data in the RAID Controller’s cache during a power failure.

8. Every RAID Controller can support only a certain number of disks, either in the same array or in multiple arrays.

9. RAID Controllers generally have a certain amount of Cache memory to improve the performance by storing and enabling access to frequently required data/ info.

10. A GUI based interface is generally available for the user to configure and maintain the Raid Controller, create and maintain RAID arrays, get notified of individual disk failures, etc.

11. Some RAID Controllers can also issue an audible alarm (like a beeping sound, etc) in case of disk failures/ new disk installations.

12. Additional (limited number of) disk drives can be added to a RAID array with certain RAID Controllers.

13. RAID levels can be optionally and selectively migrated from one level to other supported levels (For example, RAID 1 can be migrated to RAID 10), if required.

** Please note that most of these features are applicable for Computer Servers and a few of them maybe applicable to individual computers as well.

excITingIP.com

You could stay up to date on the various computer networking / enterprise IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

What is Data De-duplication?

Simply put, Data De-duplication is the process of removing duplicate data from storage systems. Consider this scenario: You are the technical manager of a IT services company. There is a huge installation file, which you send to 10 of your subordinates. They promptly download a copy to their computers. Now there are 11 copies of the exact same data both on the computers and the email server.

When a back up is taken for the day / continuously (using Continuous Data Protection or similar technologies) all the 11 copies of the same document are stored/ backed-up individually (within their respective directories), causing inefficient usage of storage resources. Instead, if the storage system can recognize that all these files are exactly similar, it can just maintain one full copy and have pointers to this copy at all other places.

So, Data De-duplication is a technology that can actively look for patterns of matching data from all the incoming data and actively replace redundant data with pointers to an original source that is saved just once. This saves a lot of disk space and also the bandwidth required for Off-site back-up / Disaster Recovery scenarios.

Data De-duplication can be done at file level (or) block level. It is often used along with compression technologies to save even more disk space. Mostly, this is a transparent technology used by the storage system without requiring any user intervention. This can be done as the data is arriving (in-line) or it can also be done as a post-processing activity.

Types of Data De-duplication:

There are two major types of data de-duplication – Block level & File level.

1. Block level Data De-duplication: All the data entering the storage system irrespective of the file size, are divided into small equi-sized blocks (few kb each). These blocks are subject to a hashing algorithm and the result of the same is stored in a separate index file, for each block. All the incoming  data are divided into blocks and are subject to the same hashing algorithm to be compared with the previously stored hashed results (of existing blocks) in the index file to determine if there are any exact matches.

If the results match, the incoming block is removed and a pointer is stored in its place referencing the hash value of the similar block in the index file. If the results do not match with any existing values, the block is stored in full and it is subject to the hashing algorithm whose result is stored as a new entry in the index file.

This technique saves more disk space than file level De-duplication because data is divided in to small blocks and where ever there is a repetition of these blocks, they are replaced by pointers. The probability of finding similar blocks is higher with smaller reference values (blocks).

Also, if there is a change in the original file, only those blocks that are affected by the change are re-written, instead of re-writing the whole file. Some vendors use variable sized blocks, instead of fixed blocks for obtaining better storage efficiency with larger files.

But, the higher efficiency of block level De-duplication comes at a cost – More processing resources, More RAM (primary memory) utilization & More Storage space (Secondary memory, for storing index files).

2. File Level Data De-duplication: The process of file level Data de-duplication is similar to block level Data de-duplication, but instead of  dividing data into small blocks and comparing these small blocks against incoming data, whole files are compared with incoming files. So, if the same file requires to be stored once again on a storage system, it will store a pointer to the location of the initial file copy instead of storing whole files again and again at different places.

The amount of storage space saved due to this technique may be lesser than block level process, but it is still significant. It also comes with faster operation, lower overheads (processing capacity, RAM capacity, etc) and lower cost.

Data Compression:

To achieve even more disk space savings, Data Compression is often used along with Data De-duplication so that data can be compressed using various compression algorithms before they are actually saved to the disk. The compression efficiency depends on the type of compression algorithm used and often results in a 2:1 disk space reduction.

Of course while retrieving the stored data, reverse processes of Data De-duplication & Data compression needs to be applied, which results in higher processing overhead. This is also true for transferring data between multi vendor storage/ back-up devices that do not recognize the De-duplication techniques used by the other vendor.

excITingIP.com

You could stay up to date on the various computer networking / enterprise IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

What is a File Level Storage System?

A File Level storage system is the most common storage system that we find with our hard-drives, NAS systems, etc. In this type of storage, the storage disk is configured with a particular protocol (Like NFS, etc) and files are stored and accessed from it as such, in bulk.

Advantages of File Level Storage System:

• File level storage system is simple to implement and simple to use.
• It stores files and folders and is visible as such, to both the systems storing the files and the systems accessing it.
• File level storage systems are generally inexpensive, when compared to block level storage systems.
• File level storage systems are more popular with NAS based storage systems – Network Attached Storage.
• They can be configured with common file level protocols like NTFS (Windows), NFS (Linux), etc.
• File level storage systems are well suited for bulk file storage.
• The file level storage device itself can generally handle operations like access control, integration with corporate directories, etc.

What is a Block level storage system?

In Block level storage systems, raw blocks (storage volumes) are created and each block can be controlled like an individual hard drive. Generally, these blocks are controlled by the Server based Operating Systems. Each block/ storage volume can be individually formatted with the required file system.

Advantages of Block level storage systems:

• Block level storage systems offer a better performance/ speed than file level storage systems.
• Each block / storage volume can be treated as an independent disk drive and are controlled by external Server OS.
• Each block / storage volume can be formatted with the file system required by the application (NFS / NTFS / SMB , etc).
• Block level storage systems are very popular with SAN – Storage Area Networks.
• Block level storage systems are more reliable, and their transport systems are very efficient.
• Block level storage can be used to store files and also provide the storage required for special applications like Databases, VMFS (Virtual Machine File Systems), etc.
• They can support external boot-up of the systems connected to them.

excITingIP.com

You could stay up to date on the various computer networking/ enterprise IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

But first let us consider the advantages of External Hard Disk/Drive:

1. Low cost.
2. Portable – Can be carried anywhere.
3. Sufficient storage capacity – 320 GB to 1.5 TB in popular models.
4. Automatic backup software.
5. Encryption software.
6. Connects to any Computer/Laptop using the USB port/ eSATA port.
7. Available anywhere and simple to use.

Now, let us look at the advantages of NAS box:

1. Connects to the Network directly using RJ-45 Ports.
2. Multiple users can simultaneously connect to, store & share files using the NAS box.
3. Cost is almost equivalent / slightly higher than an external Hard-disk Drive of similar capacity.
4. Automatic back-up/ encryption.
5. The contents of the NAS box can be remotely accessed (If it supports DDNS) from anywhere over the Internet.
6. Media stored in the NAS box (audio/video) can be streamed within a small network, without users having to download files every time.
7. NAS box enables companies to create common storage and backup mechanism so that the data need not be duplicated in all the individual users PC/ Laptop.
8. Some NAS Boxes support multiple disk-drives to be mounted on the same box for additional storage capacity.
9. Some NAS Boxes support RAID with multiple disk-drives to ensure that the data is available even if one of the disk-drive fails (Redundancy).
10. Some NAS Boxes support Bare Metal Recovery (Restoring the state of the NAS box to how it was, on a certain date/time).
11. NAS a higher storage capacity – 1 TB/ 2 TB/ 3 TB/ 6 TB, etc.
12. Independent device – Can be Always ON and not dependent on a particular user’s laptop. Anyone connected to the network can access it anytime.
13. Even if you just have a Wi-Fi network, you can connect the NAS box to one of the Wired ports of the broadband router and it becomes accessible to everyone over the Wi-Fi Network.

So, if you are buying a storage/ backup device for your SOHO/ SMB or even your home, why not consider a NAS box?

excITingIP.com

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

RAID Systems (In Computers, Servers, NAS, etc) use multiple disk-drives to store data to prevent data loss when individual drives fail & to increase the read/write memory performance. Software based RAID systems provided by the operating systems use the system CPU to create and manage RAID drives while Hardware based RAID systems require a dedicated hardware RAID controller/ processor, that can access multiple disks simultaneously for the same purpose. There are many types of RAID systems and the common ones are RAID 0, 1, 5, 10, etc.

In RAID 0, data is split and written on two disks – Half of it on one disk, and the other half on the other. This type of RAID system is mainly used to improve the read/write operations & memory performance in general as it employs Data striping with multiple disks. It offers no protection for data, as such.

In RAID 1, data is written on one disk and simultaneously copied to another disk (Disk Mirroring). Even if one disk fails, there would be no disruption of services as the system can use the other disk which contains the exact copy of the failed disk. But in RAID 1, the usable disk space is 50% of the total installed capacity.

RAID 10 is a combination of RAID 1 & RAID 0. It needs a minimum of four disks to configure. In this, two disks are paired as one set and data is mirrored within each set, which accounts for data recovery. Data is still split into two and written alternatively on each set of drives – one by one which accounts for data striping and better performance. Usable disk space is 50% of the total installed capacity.

RAID 5 requires minimum 3 disks to implement and it uses parity. If a disk fails, using the parity value (which is a function of data on other two disks) the data which was stored on the failed disk can be calculated for each block. RAID 5 gives good performance as it can use data striping to write across multiple disks & it can also give a better storage efficiency – Usable capacity: Total Capacity Minus Disk capacity of one disk.

Though RAID systems provide for disk redundancy (when one of the disks fail), there are situations where data might still be lost/ inaccessible. Can data from such RAID systems be recovered?

Some reasons for data loss in RAID systems:

1. More than one drive failing at the same time / another drive failing just before the faulty drive is replaced.

2. RAID Controller Card failure/ RAID configuration gets corrupted.

3. Accidental formatting of all Disks/ Bad Sectors.

4. Accidental file deletion (manual errors).

5. Recovery failure after a failed drive is replaced, etc.

Data Recovery for RAID Systems:

There are multiple methods used in recovering data from failed RAID systems, and some of them are discussed below.

A method called ‘striping’ is used in most of the RAID systems which stores the data to various drives in small blocks of 128/256 Kbps (for example). This is referred to as stripe size. So, the files that are smaller than the stripe size (like small text documents, spreadsheets, etc) can be recovered in full as they can be accommodated into a single block and hence may not be split.

But even larger documents that are split into multiple blocks and stored in multiple drives can be recovered by a process called ‘De-Striping’, where each block of data is copied sector by sector from all the drives involved into a new destination and attempts are made to arrange all the sectors in the right order to reconstruct the (larger) original file. It may not be possible to recover all the files like this but depending on the extent of damage, most of them might be recovered.

If the drives in a RAID array are accidentally formatted, it might still be possible to recover data from them depending on how they were formatted – If an OS/ Software was used to format the drives (in high level), the old data might still be available in the free space and can be recovered. But if a Controller was used to format the drives (in low level), there is a chance that data might have been totally erased.

There maybe cases where some hardware components of RAID systems might have been physically damaged – In such cases, the failed components may be replaced to recover the data.

It might be a good idea to take professional help from companies specialized in RAID recovery in case of RAID system data loss/ failure.

excITingIP.com

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

What is a Self Encrypting Drive and Why is it used?

Encryption, has long been used to protect confidential documents and communications. Software based encryption tools are available to either selectively encrypt certain files (or) encrypt whole disks. But software based encryption tools mostly leave the decision of encrypting (or) not, to employees. So, there is a chance that employees may not encrypt important information stored in their laptop disks (or) encrypt only those files which they consider as important. So, that’s a bit like leaving it to chance.

Encryption, as a technology has matured enough and these days encryption is done using 128 bit/ 256 bit keys. Encryption can be done using existing hardware / processing resources and is very difficult to break (It might take many years for hackers to decrypt info using brute force methods). But authentication, is as important as encryption. Even the biggest encryption key may be useless if the password is ’123′.

Self encrypting hard drives are transparent to users. It doesn’t matter what operating system / applications are being used, as the encryption is done using specialized hardware ASIC modules on the disk itself. This improves the encryption performance and precious CPU cycles can be utilized for other processes.

Some vendors support pre-boot authentication, which allows the user to enter a password to decrypt the disk contents, before the booting process starts. Its possible to integrate self encrypting disks to Windows single log-on / password update policy in order to support a single password log-in and subject it to corporate policies like mandatory password changes – once in 90 days, for example.

Some vendors even offer centralized administration features like management of users, user credentials (passwords), access rights, resetting forgotten passwords, etc. Its possible to integrate some self encrypting drives with corporate directories like Active Directory. Other enterprise systems like Servers, NAS/ SAN Storage etc can also use self encrypted disks (perhaps in future, if not already). Even Solid State Drives (SSD) manufacturers are coming out with Self encrypting solid state drives.

The main advantage of encrypting the whole disk is to secure confidential data/ information from being exposed if a laptop/ disk drive is stolen. Think about this – even if you give a laptop for service/ disk for replacement, some sensitive information can be recovered/ stolen. In an auto-lock mode, disks are by default kept encrypted and even if a colleague tries to break into your laptop and look for/ steal some data, they may not be able to. And besides, it is easy to dispose off the self encrypting drives as the encryption key can be just removed from the drive in order to render the disk useless.

Speaking of keys, the encryption keys are stored in a secure/ inaccessible place in the disk along with other manufacturer information. This is the key that is checked against user-input credential which should generate an identical copy of the key, for the authentication to succeed.

Software Encryption (vs) Self Encrypting Drives:

There are software tools that are installed over an operating system to either selectively or completely encrypt the disk drives. But these are operating system specific, unlike self encrypting drives that are operating system independent.

Software based encryption has some advantages like encrypting only the important/ required files and folders, encrypting email attachments, can be used along with digital signatures, encrypting external hard disks/ USB drives, encrypting content stored in mobile / tablet drives, etc.

But self encrypting drives use hardware based ASIC chips to encrypt all the content stored on disks. They don’t use the main CPU and hence  encryption performance is faster and better (Often encryption is transparent to users). The keys are stored in a secure (inaccessible location) on the disk drive itself and most of the operations that can be done using normal drives can be done by self encrypting drives without introducing major changes – For example, once decrypted, RAID controllers work in the same way for both.

excITingIP.com

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’.

In the above diagram, both the traditional set up with FC HBA + NIC (and) the new set up with CNA/FCOE are represented. As you can see, with the set up in the first diagram, two separate adapters are required on the server to connect to Ethernet based Computer Network and FC based Storage Network respectively.

But the set up in the second diagram requires just one adapter (Converged Network Adapter – CNA) which carries both Ethernet traffic as well as FCOE traffic in a single cable. This cable connects to one of the Ethernet ports in the Converged Switch that has both Ethernet as well as Fiber Channel ports. This Converged Switch converts the FCOE traffic in to Fiber Channel traffic to be sent to the FC SAN over the Fiber Channel Network. The computer network traffic is directly sent to the LAN over the Ethernet Network.

Salient points about CNA – Converged Network Adapter:

• Fiber Channel Over Ethernet (FCOE) is a standard that encapsulates Fiber Channel (FC) traffic over the Ethernet packets to be run on Ethernet Networks.
• The Converged Network Adapter (CNA) connects to the server using the PCI Express expansion interface.
• CNA can offload the FCOE protocol processing task, thereby freeing the server CPU resources from doing that task.
• The new setup works with Fiber Channel Networks, FC Switches and Fiber Channel Management Utilities. Some vendors even offer converged management utilities for their Fiber Channel products along with the CNA – Converged Network Adapters/ FCOE.
• The main advantage of Converged Network Adapter is the reduction of the number of adapters required on the server, reduction of the amount of cables, reduction in the amount of switch ports, reduction in the number of PCI Express slots (on the server).
• Since most of the Converged Network Adapters support 10 GBEE (10 Gigabit Enhanced Ethernet), the performance of the network carrying both network and storage traffic should be good enough. And since CNA follows the same Ethernet standards, 40 GBEE & 100 GBEE will be supported in future.
• There are special switches that support FCOE capabilities on their Ethernet ports. They either come as stand alone switches with both Ethernet/FCOE Ports & FC Ports (or) Chassis based switches that support both Ethernet/FCOE & FC Blade modules.
• Some vendors offer these FCOE supported Switches as Ethernet only (with a lower price), to connect to the computer network presently.  But, FCOE & FC functionalities can be included in the future (through a software upgrade), whenever required.
• Converged Enhanced Ethernet Standard for Data Center Networks is recommended for carrying FCOE traffic, but is not mandatory. So, FCOE based switches can work with normal Ethernet based switches as well.
• CEE – Converged Enhanced Ethernet makes FCOE Storage Network more reliable as it eliminates the lossy behavior of Ethernet by introducing some features like Priority flow control, Congestion Notification, Enhanced Transmission Selection, DCB Capability Exchange protocol, etc.
• CNA supports Copper and Fiber Transceivers with SFP Plus Form Factor.
• Many types of cables like Copper Twin-ax, Copper Twisted Pair (Cat 6a), Single Mode & Multi Mode Fiber are supported by CNA/FCOE.
• Sever Virtualization Vendors support Converged Network Adapters for further reduction in the required hardware for Data Centers.
• Converged Network Adapters can be used as stand-alone 10 GE NIC (Network Interface Cards), if storage networking is not immediately required. FCOE & FC SAN functionalities can be added later on.

excITingIP.com

You could stay up to date on the various computer networking technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

The first image defines what is iSCSI and what exactly its role is, in Storage Area Networks. Convergence is happening over IP everywhere (one reason why IP is exciting!), and even the Storage Area Network is expected to converge over IP networks through iSCSI especially with the anticipation of lossless Converged Enhanced Ethernet. The other popular convergence option being FCOE (Fiber Channel Over Ethernet).

The second image lists the advantages of Storage Area Networks, in general. If you have been hesitating to implement/ upgrade to a Storage Area Network (SAN) because of cost, complexity, or any other reason, iSCSI SAN now offers you lot of advantages in addition to the ones mentioned above. Let us look at some of them:

• With iSCSI, a separate network for SAN is not required as it uses the existing IP Networks and components (NIC, Switches, Cables etc) to create the Storage Area Network.
• The cost of creating a iSCSI SAN is very less when compared to the cost of creating a FC(Fiber Channel) SAN.
• iSCSI based SAN can co-exist with the existing FC based SAN. Customers have the option of retaining their existing investments on FC SAN’s and still expand by adding additional storage capacity using iSCSI SAN.
• iSCSI SAN does not have any distance limitation. You can have a Data Center anywhere in the world and still back up/ restore the data remotely (over WAN/Internet) to your Local Area Network/ vice versa.
• Since iSCSI SAN can be located anywhere, they are very useful for disaster recovery objectives.
• iSCSI SAN can use either a specialized HBA (Host Bus Adapter) to connect the servers to the SAN (or) just use the standard NIC Cards/ Ethernet ports for the same. This enables server I/O consolidation and reduces complexity/ cost.
• Gigabit Server Adapters (NIC Cards) are normally available in the market, and iSCSI can use them to connect to the network at Gigabit speeds today. They are available in Single Port/ Dual Port/ Quad Port etc, and connect to the server using PCI Express slots. Very soon, even 10 Gigabit Ethernet NIC Cards are expected to become popular, which offers much more capacity/ speed for SAN performance.
• Since iSCSI uses the normal IP based network components, its easy to learn, implement and maintain. Contrast this with Fiber Channel (FC) based SAN which requires a high level of expertise to create and maintain them.
• iSCSI is very much suitable for implementation of SAN in virtual server environments as it supports software initiators that make such integration easier. Also, since iSCSI supports whatever bandwidths supported by IP Networks, it can support 10 GE which might be required for virtual server environments.
• iSCSI allows direct backup’s to tape or disks, even from certain virtual servers.

Limitations / Disadvantages of iSCSI SAN:

• The IP network, currently is a best effort network. Hence it may drop packets or deliver them out of order during network congestion. But these attributes are not favorable for storage area networks as they need to be highly reliable. So, till the lossless Converged Enhanced Ethernet technologies are standardized and implemented, iSCSI SAN may still not be the most reliable option when compared to FC SAN.
• Server CPU’s may be burdened with iSCSI and TCP/IP Stack processing if HBA/ NIC cannot offload that function. This may result in the server processor cycles not working to their capacity for application processing, which is what they are primarily supposed to do.
• Since iSCSI is generally not run on a separate network (but is mixed with the IP network traffic), there may be networks congestion/ bandwidth constraints during peak hours especially if the network is not designed to support the bandwidth required for both the applications.
• iSCSI operates on clear text protocol (normally) and hence there is a chance that it might be exposed to hackers who are more familiar with the widely used IP Networks.

Salient points/ Good practices for iSCSI SAN that might be useful:

• CHAP (Challenge Handshake Authentication Protocol) can be provisioned for authentication of iSCSI messages for security. Encryption, even if possible, may not be feasible – currently.
• Some specialized HBA’s/ NIC Cards can offload processor hungry processes like iSCSI and TCP/ IP Stack processing. This can make the server processor work more efficiently.
• Major Operating Systems (Including Windows and Linux) provide built-in software initiator drivers for iSCSI which makes their integration with IP network components, easy.
• Its better to have a dedicated port for iSCSI SAN connection from a server (instead of sharing one network port). Better still, redundant ports can be provided for High Availability/ Link Aggregation.
• Its a good practice to have iSCSI packets flow on a separate VLAN so that the storage traffic can be logically separated from the general network traffic. Better still, if iSCSI can be operated on its own physical network segment (with dedicated switches/ cables, etc).

excITingIP.com

In case you have any points to add on the above topic, you could comment below. You can also contact us using the contact form. In order to be updated on the various computer networking technologies, you can subscribe to this blog using your email address in the sidebar box mentioned as “Get Email Updates When New Articles are Published”.

What is the need for FCOE – Fiber Channel Over Ethernet?

High density storage for large organizations use SCSI and FC – Fiber Channel protocols. So, they reside in a separate network with their own adapters (HBA), FC Ports, FC Switches, Cables, etc. Fiber Channel has been reliable (lossless), and has good performance, speeds and hence they are very popular for large scale storage systems.

But, with our ever pervasive IP Networking protocol standardizing on the Ethernet medium, all companies have their IP Networks running over Ethernet. People saw the opportunity to converge both the Computer Network and FC SAN Network, and hence two major protocols were born to enable such convergence – iSCSI and FCOE.

iSCSI (SCSI over TCP/IP) replaced the FC stack with the TCP/IP stack so that the storage traffic could be transported over the low cost TCP/ IP Networks. This works well enough for small/ medium scale implementations but large networks still demanded the reliability of the Fiber Channel (FC) Network. Hence, FCOE – Fiber Channel Over Ethernet was conceived.

What is FCOE – Fiber Channel Over Ethernet?

FCOE or Fiber Channel Over Ethernet transports the SCSI storage data (used in Fiber Channel – FC networks) using Fiber Channel Protocol stack instead of TCP/IP stack; using the Ethernet infrastructure (NIC, Cables, Switches, etc). FCOE maps the FC commands and data directly into Ethernet frames (Fiber Channel frames are encapsulated in Ethernet frames) to create FCOE and the mapping is 1:1 meaning, there is no segmentation or compression of FC frames.

But, as we know, Ethernet is a lossy medium – it provides a single best effort pipe that drops packets during a network congestion. So, in FCOE, FC is simply encapsulated and run over a lossless Ethernet infrastructure.

So, how is that Lossless Ethernet Infrastructure created by FCOE?

FCOE has to create a lossless Ethernet environment in order to ensure the reliability of large scale storage data transportation. Two standards enable that – The Data Center Bridging (DCB) and Converged Enhanced Ethernet (CEE). A few of the enhancements to Ethernet (to make it lossless) are described below to get an idea about DCB and CEE.

IEEE 802.1Qbb: Priority Flow Control - IEEE had earlier defined the means to categorize traffic according to its priority – IEEE 802.1p (QoS Standard). So, the new standard IEEE 802.1bb takes advantage of this standard by first allowing to classify the traffic in to eight categories (lanes), each of which could be assigned a priority level. Priority Flow Control issues a ‘Pause’ command that allows to halt FCOE traffic during congestion so that the losses can be minimized and it uses the priority level to recognize FCOE from other types of traffic. So, the administrators can create lossless (virtual) lanes for FCOE traffic and lossy (virtual) lanes for normal IP based traffic.

IEEE 802.1Qau: Congestion Notification - Congestion is measured at the congestion point, where ever it is happening but the action (like rate limiting) is taken at the reaction point (originating point). For example, an aggregation switch can ask an edge switch to stop (or limit) its traffic from a particular port, if congestion is encountered, through this standard.

IEEE 802.1Qaz: Enhanced Transmission Selection - High Priority traffic (like FCOE) can be allocated with a minimum guaranteed bandwidth so that all the other traffic on the network does not congest the path due to their high volumes.  However, if FCOE traffic does not fully utilize the path (its reserved capacity), then the bandwidth is  allowed to be used by other types of traffic and this can be controlled dynamically by this protocol.

What are the components of FCOE – Fiber Channel Over Ethernet Infrastructure?

There are basically three components of FCOE infrastructure:

1. Converged Network Adapter (CNA) – The CNA is a single adapter in the server (that attaches to PCI Express Slot) which can provide the functionalities of both Ethernet NIC (Network Interface Cards) and FC HBA (Host Bus Adapters) virtually. That means, the server still sees two Interfaces and it sends the IP traffic to the NIC and FC traffic to the HBA. But the CNA collects both of them, and transports it over a single Ethernet cable after having wrapped all the Fiber Channel (FC) frames to Ethernet frames.

2. FCOE Links – The FCOE infrastructure utilizes the same Ethernet infrastructure used by the TCP/IP Network. So, it uses the UTP Copper Cables, Optical Fiber Cables and even the low cost Twinax cables that use the SFP+ Interface to carry 10 GE for short distances.

3. FCOE Switches/ Network Switches supporting FCOE protocol – The Fiber Channel SAN’s understand only the Fiber Channel Protocol and Fiber Channel Interfaces. So, there needs to be an intermediary that separates the FCOE traffic from the regular IP traffic and connects to the FC SAN’s directly. This intermediate functionality is given by FCOE Switches (or) Network switches (with FC Ports) supporting FCOE protocol. So, the HBA’s from the Servers connect to this FCOE Switch which in-turn connects to the SAN Network using FC ports and IP network using IP Ports.

What are the advantages of FCOE – Fiber Channel Over Ethernet?

• FCOE reduces the two network adapters (HBA for storage connectivity, NIC for computer network connectivity) and two individual cables from each server to just one, thereby simplifying the network.
• FCOE can carry traffic over the Ethernet medium and hence use the familiar and easily available copper UTP cables and Optical Fiber Cables.
• One Network Adapter instead of two gives some power savings for the server.
• There are certain I/O Virtualization solutions that support FCOE which reduces the total number of server adapters (CNA) for a group of servers by consolidating them on to a I/O Virtualization appliance and allowing the servers to share the common pool of adapters. This, reduces the total number of CNA’s required for a group of servers. The servers themselves connect to the I/O Virtualization appliance through interfaces like PCI Express / appropriate cables from there. But certain proprietary vendor based drivers may have to be installed, to complete this setup.
• The performance of FCOE network is is comparable to that of FC/ IP networks with FCOE supporting the speeds of the Ethernet network ( Up to 1/ 10 Gbps currently and expanding to 40 Gbps and 100 Gbps in the future).
• FCOE can be used in virtualized environment (Server Virtualization) and is quite advantageous to use in such circumstances.
• FCOE, unlike iSCSI, is a very reliable Storage transportation protocol, and can scale up to thousands of servers.
• Since FCOE just encapsulates the FC data onto Ethernet frames for transportation only, all the existing administration tools and workflows for FC (Fiber Channel) remain intact and hence the investments in existing Fiber Channel storage is preserved while the reliability of FC is also maintained.
• The support from Network Switch vendors for FCOE by offering converged switches with both Ethernet and Fiber Channel ports strengthens the case of FCOE.

Dis-advantages / Limitations of FCOE – Fiber Channel Over Ethernet:

• The only Ethernet component that is currently compatible with FCOE is the cables. Everything else will have to change for implementing FCOE. So, the actual component saving today would only be the amount and cost of cables!
• The cost of a Unified CNA (though it is coming down) might still be more than the cost of the HBA and NIC combined!
• FCOE is currently restricted to Access Networks only (Server to Switch Connections). The distribution and core storage networks are still in Fiber Channel and will continue to be in Fiber Channel till FCOE technology matures enough to create its own FCOE SAN networks.
• iSCSI proponents might still argue that changing one disparate network in to another does not much amount to convergence of storage and network infrastructures.
• Security on FCOE networks might have to be re-looked as the network is now running over Ethernet and hence is more easily accessible than their Fiber Channel counterparts.

excITingIP.com

In case you want to add any more points or have clarifications, you can use the comment form below to drop a comment on this article or contact us using the contact form. You could keep yourself updated on the various Computer Networking technologies by subscribing to this blog by submitting your email address in the box titled “Get Email Updates When New Articles are Published”.

Before exploring Bare Metal Recovery, we should see how a normal back-up process works briefly- You have a back-up agent installed on the clients/ servers that need to be backed up and a back-up server/ appliance that actually coordinates and stores the backed up data. Of course, the main aim of back up is to restore the data and hence the restoration process (on a machine whose hard disk was replaced, for example) basically consists of reinstalling the operating system first, then the applications and then the data and then the settings/ configurations. All these are individual processes that needs to be done individually.

As you can guess, the chances of making errors in the above process are higher, as chances of human intervention (and reconfiguration) going wrong is more during the restore. It also takes more time to properly restore and check if everything has been done correctly, and troubleshoot if not. Bare Metal Recovery was introduced to make this process easier, better, faster and fail-proof.

What is Bare Metal Recovery/Restore (BMR)?

If a server or a group of servers are totally destroyed (by a disaster like flood, fire etc), Bare Metal Recovery still enables the IT team to recreate them from scratch (on a different system/disks) including the Operating System, its configuration/ settings, patches, applications running on the operating system and (perhaps with the help of a back up software) even the data! The BMR solutions mostly follow automated processes with minimal human intervention – hence they are fast and mis-configurations could be avoided. So, the hardware on to which the system is being recovered can be devoid of any operating system, file formats, partitions etc, and the BMR would recover all of them! So, the restore disk could just be a bare metal (roughly)!

Some BMR solutions take snapshots (images) of the entire disk and back it up frequently and restore to the most recent snapshot. This however makes it difficult to restore the data in a dissimilar hardware configuration. Some more vendors tweak their backup solution a little and add the BMR component to the same where each component(operating system, application, etc) could be separately installed (with changes, if required) as the backup solution stores the data and BMR solution backs up the operating system and the settings/ configuration.

What are the advantages of Bare Metal Recovery (BMR)?

• BMR is faster than manual backup processes.
• BMR is prone to lesser errors as human intervention in the backup process is minimized.
• Saves a lot of administrative efforts during a disaster as all the settings/ configuration remains intact and the trouble of e-installing  individual softwares , operating systems manually can be avoided.
• Certain BMR solutions allow incremental backup and restoration, which saves a lot of storage space and bandwidth, and helps backing up to a remote location.
• Some vendors offer integrated BMR and Backup solutions which enable BMR to be added as an additional component of the regular backup where the configuration/ settings/ operating system etc are restored from BMR and data is restored from the normal backup software.
• BMR can restore not only to the latest backup point, but also to any available (earlier) backup points.
• BMR can recover the whole server, operating system and applications (or) selectively recover some of them.
• BMR can be used for automating server migrations (to other physical or virtual infrastructures) too, as the process is almost the same.

What are the disadvantages of Bare Metal Recovery (BMR)?

• A common problem in BMR is the recovery to dissimilar hardware configuration (when compared to the original systems from which they were backed up), as the device drivers for the new hardware will most probably be different. But certain BMR solutions can handle this and recover to any new configuration of the hardware.
• BMR solutions are mostly specific to the operating system of the systems that need to be backed up and restored.
• BMR solutions generally do not support recovery from tape drives and even the ones that do, require that the BMR software and the tape drives be purchased from the same vendor.
• While backing up servers are critical, a lot of data is resident on mobile devices like laptops and tablets. More over, they keep frequently changing and sometimes cannot wait till the scheduled backup happens. In those cases, Continuous Data Protection might be more apt than bare metal recovery.
• BMR is more focused on disaster recovery and a company that has no investments in a backup software, would hesitate to invest for pure DR. Even if the company has a backup solution in place, there is no other option than to buy the BMR solution from the same vendor. And some backup solution vendors don’t integrate with BMR.

excITingIP.com

In case you have any point to add or have any questions, you can use the comment form below or contact us using the contact form. You could keep up to date with the latest technologies in the computer networking industry by subscribing with your email address in the box titled “Get email updates when new articles are published”

]]> http://www.excitingip.com/934/bare-metal-restore-recovery-backup-for-disaster-recovery/feed/ 1