We provide storage. As a simple explanation, it's a bit like your hard drive on your laptop or PC. The difference is, it is usually connected to a network (like your intranet or the internet).
- You can store applications, documents, live operating systems (like Windows).
- You can track changes and you can back it up so you will never lose your data.
- You can manage your own storage, or we can manage it for you.
- You can organise your storage so that even if your business burns down, if you have another office - you can use the storage in the other office immediately so you won't even notice if something really bad happens (this is a type of redundant system used in a disaster recovery scenario).
- Even if you don't have another office - you can use our facilities.
- It is more than just "saving documents automatically"
- You can expect to be able to transfer your entire business over to another location (see business continuity) if something ever goes wrong.
- Additionally you can keep everything backed every moment of every day (which is mirroring), or you can choose to backup when you want (like a specific time).
- Your administrator can recover documents for you or you can manage the recovery yourself (every staff member can do this if required).
The storage we sell keeps your organisation up and running. Always.
It is not a technical issue to implement backup and recovery in your organisation. It's a business imperative. Good storage is what is required for your data.
Is data the backbone of your business? Many organisations don't realise the dependence they have on their I.T. systems and day-to-day data. Data includes the files you actually work on, the machines you use, the operating systems and applications that help you run and manage your business.
You may require adequate protection of your data.
-
Backup. Archive documents and servers
-
Redundancy. Restore and recover your backups (instantly)
- Reliability
- Resilience
- High Availability (zero downtime)
-
Store. Storage that delivers maximum performance and scalability
-
Encrypt. Secure your data across your enterprise.
-
Growth. Manage explosive data growth in a controlled manner.
-
Productivity. Use the same framework for backups and active systems.
-
Collaborate. Improve compatibility between systems. Making everything do more.
-
Unify. Discard or unify systems and resources so that you can do more with what you have.
-
Improve business processes
-
Adapt. Adapt and scale infrastructure
-
Supported. 24x7 global support
-
Compliance. Easily implement data security and compliance requirements.
-
Integrate. Integrate into your existing applications.
-
Reduce. Single enterprise-wide operating costs
If your organisation relies on information it will also need a useful recovery plan. Our business continuity consultants can help.
Direct Attached Storage
Direct Attached Storage (DAS) is storage that is directly connected to your operating/application hardware. Direct attached storage has become antiquated due to its disparate nature. It is hard to share, it is hard to scale. A simple example would be a hard-drive in your PC.
Network Storage
Network storage is as it sounds available to the network. This allows you to share the storage across large distances with many users. Depending on the architecture of network storage, it scales easily.
NAS (file oriented)
Network attached storage (NAS) allows you access to files. The protocol controls access permissions, authorization and automated file level locking. Examples of protocols are SMB/CIFS (Windows), NFS (Unix, *nix), DAFS (Unix, *nix).
SAN (block oriented)
Storage Area Network (SAN) allows you access to a block level device via your network (they appear locally connected). This allows you to create or mount remote hard-disks over the network. The SAN topology is suitable for scalability and dynamic, fast recovery (including disaster recovery). An example of a SAN would be an iSCSI target hosting your database data on a storage network appliance which you connect to on your local area network (LAN).
Virtualised
“Users need block access and file access – what they really want is a single storage platform that does both.” (Steve Duplessie, Enterprise Storage Group, http://www.enterprisestrategygroup.com )
Any good network storage will allow you to access it on either the block or file level. Virtualised network storage abstracts its data so any readily accessible protocol can utilise it. This way you can use it to share files or disks (at a block level). NetApp the innovator of the first multi-protocol appliance (1996) calls this FAS (fabric attached storage).
You can choose to deploy your storage across multiple sites. The primary advantage for doing this is to ensure business continuance and all of your data is protected.
As we only recommend network storage, irrespective of your deployment method it should be accessible from anywhere. There are many software applications available to enable your staff to remotely manage and check-in their data.
Storage (for organisations) is typically provided by appliances. Appliances are dedicated machines that intelligently manage your data. Depending on the appliance, it can provide
-
snapshots (high performance point-in-time copies)
-
mirrors (live replication)
-
clones (accessible replicas)
-
maximise storage utilisation (thin provisioned)
-
security
-
stability
-
recover rapidly
-
data-protection
-
consistency
-
simplicity
-
ongoing protection
-
hardware compatibility
-
software compatibility
-
network compatibility
-
protocol compatibility
-
permanence, especially for regulated and reference of unstructured data
-
support
-
tape-compatible
A multi-site deployment utilises more than a single appliance. The advantages are:
-
allows you to distribute your data worldwide
-
you can replace (offline) tapes with low-cost rapidly accessible near-line storage
-
overcome DAS bottlenecks.
-
achieving "near-perfect" database availability.
-
economical support for long-term growth and innovation
-
fast failover with synchronous replication for rapid recovery and business continuance (disaster recovery over distances commensurate with dense wavelength division multiplexing (DWDM) limits)
Single-site Deployment
Single-site deployments can take advantage of all the benefits of a good appliance. The only further requisite is that you must ensure business continuity by some other means. You may wish to leverage our existing technology and outsource your data with us to get the added benefits of having a multi-site deployment.
The choices (see here for a simpler explanation).
Nearly every business requires redundancy, performance and reliability of its data to ensure high and consistent service of its clients. We can provide you with innovative storage solutions that will cater for all your requirements (see here for a simpler explanation).
Primary Storage
We recommend you use network storage (SAN/NAS) and a dedicated appliance for critical business data for the following reasons:
-
Resilient. Appliances provide automated self checking and constant monitoring of data (or at least they should). This ensures your data does not corrupt.
-
Scalable. Appliances give you the ability to add extra disk as your business grows. If you grow too big for the appliance you purchase, you should be able to seamlessly upgrade to a newer one – or quickly and easily access older data (a near-line Virtual Tape Library (VTL) might be a good idea).
-
Compatible. Appliances allow you to communicate with all the computers in your network using any operating system or major protocol at the same time (or at least they should).
-
Consolidation. An appliance will allow you to consolidate unwanted servers. A good appliance will allow you to consolidate your other appliances (even if they're manufactured by other companies).
-
Recovery. An appliance allows everyone to go back to a point in time...
-
Redundancy. It should be quick and easy to switch to an appliance that fails due to:
-
Easy. The appliance should be easy enough to use that someone with limited I.T. experience can manage it.
-
Supported. An appliance is supported. The device should let you know how healthy it is; and if something ever goes wrong with it, it should automatically request its parts to be replaced.
-
Speedy. It should be fast.
-
Snapshots should be instant
-
Recovery should be instant
-
Mirroring data across sites should be as fast as the network
-
Fail-over should be instant
-
Making changes to the appliance settings should be fast
-
Open. It should allow users to leverage existing technologies:
-
Backup software (Symantec NetBackup etc.)
-
Virus scanning (Appliances, or software)
-
Encryption methods (DataFort)
-
Operating system integration (Synchronise your data from within your client OS)
-
Application integration (SDK availability)
-
Tape library friendly.
-
Affordable. An appliance should be able to pay for itself in value. If you save on consolidating servers or whether it saves your business one day – the appliance will provide you tangible value.
GET A QUOTE FOR PRIMARY STORAGE
Secondary Storage
In the event of failure of your network, hardware or storage devices you'll need a backup. We recommend keeping at least a second storage appliance (as mentioned above). Other things you should consider:
-
Tape backups (or Virtual Tape Libraries)
-
Tape storage (secure offsite storage of your data especially useful for business continuity)
-
Encryption (offsite encryption of data and tapes)
-
Redundancy (Reduce single points of failure in your infrastructure: Power, power grids, telecommunications, network, servers, physical location etc.)
-
Downtime (set recovery time objectives (RTO), Maximum Tolerable Period of Disruption (MTPD), Recovery Point Objective (RPO) and provision solutions around them) . For more information see business continuity solutions....
-
Testing chosen recovery methods
-
Outsourcing your secondary storage
GET A QUOTE FOR SECONDARY STORAGE
The benefits of outsourcing your backup data are many:
-
provides affordable business continuance
-
outsource corporate governance and compliance requirements to specialists (includes regulatory compliance regarding reference material, business critical information etc.)
-
leverage existing infrastructure, minimising initial start-up costs
-
includes outsourced support & management of infrastructure
-
provision of secured yet accessible and easily downloadable copies of your data
-
large amounts of data are possible
-
backups can be (live) operating systems, file systems, databases etc.
-
you can ensure your data wont be compromised
-
scalable – we can provide you with your own off-site secondary storage when you require it - ensuring complete independence and autonomy of your data if required
-
a local repository where you can (if required in an emergency) get to your data and do a large restore (>1TB) in minutes
-
a local repository where you can download at optical fibre speeds
-
work with your data online offering complete application and data fail-over
-
can keep version history, so if you overwrite a file with a corrupted one you don't lose EVERYTHING
-
back up differentially (only backup changes)
-
can provide a high-bandwidth, centralised point for all your branches (or people working off-site)
-
can provide you with a virtual I.T. department
We expect our clients to keep a copy of their own data. We have both software and hardware solutions that will keep the data synchronised between the data off-site and on-site that will unify all your branches and team working throughout the world.
If you have interest in storing data locally in Queensland or provision an interstate business continuity solution please contact us.
If you have interest in storing data locally in New South Wales or provision an interstate business continuity solution please contact us.
NetApp products:
- FAS2xx
- FAS2xxx
- FAS3xxx
- FAS6xxx
- V-Series Storage (Consolidate existing SANs)
- Nearstore VTL (Virtual Tape Libraries)
- Decru DataFort (Storage Security Appliances)
These devices are certified for use with many other vendors hardware and software including Cisco, Brocade etc.
Innovator. Many if not all of the features in NetApp products were developed by NetApp. Many of the features have been badly copied by its competitors. Look into snapshot technologies provided by other manufacturers.
Leader. NetApp is the fastest growing storage company in the world. Outpacing the industry by 3x.
Support. NetApp provides global, award-winning support.
Simple. NetApp provides a single scalable storage platform with unmatched administrative simplicity.
Easy. High productivity management tools integrated with servers and applications.
Stable. Leading backup, restore disaster recovery and data security solutions.
Secure. Flexible archive, compliance and data security solutions with seamless encryption.
Cost-effective. Reduce cost and complexity. Consolidate.
Protected. Minimize risk & legal compliance issues.
Reliability. Reduce unavailability (with mirrors, backup,restore and other reliability solutions)
Scalable. Seamless scaling and tiered storage
Unified. Able to consolidate your existing technology and SANs
Competitive. Price competitive. Performance competitive.
-
Network Appliance FAS3070 and EMC CLARiiON CX3-80: Comparison of Performance and Usability; VeriTest an independent testing service (The full VeriTest report is available online at http://www.lionbridge.com/lionbridge/en-US/services/outsourced-testing/competitive-analysis/Netapp.htm)
-
Conclusions:
-
NetApp Delivers Better Performance and Lower Latency
-
NetApp Is Faster for a Single Application
-
NetApp Systems Enable Greater Responsiveness
-
NetApp Systems Require Less Storage
-
For maximum data availability, customers can deploy NetApp SyncMirror to achieve a level of resiliency that no other storage vendor offers. SyncMirror is local RAID mirroring between two separate volumes on the same storage system. NetApp MetroCluster is a cross facility version of SyncMirror. (The Private Lives of Disk Drives , Rajesh Sundaram , http://partners.netapp.com/go/techontap/matl/sample/0206tot_resiliency.html)
-
More comparisons:
-
Want to read more?
Snapshots - backup and recovery points in time are an important part of managing your data. They enable you to restore corrupted files or mistakes you may have made and allow you to review the changes yourself. There are many solutions available for managing versions and versioning.
As mentioned in "storage types" it is possible to store any data - files or block level data. Whether you are operating your programs off a networked disk, or storing them on a shared network location - it is possible to backup the application and take point in time copies of it.
Yes. Hosted virtual machines (onsite/offsite) is a good way to consolidate your servers and save dollars. By hosting your applications, including your operating systems and virtual machines off-site improves data resilience. You could run your entire office online.
Look into data-deduplication to save even more!
Deduplication is a technology where existing data is recycled if it is identical to some other data that already exists in your system. It has a performance deficit which is as low as 3%. An example would be a “cloned” virtual machine. If you had 1000 identical Vms in a data-centre, you could be looking at 3TB if they were not deduplicated. With deduplication you would expect 3GB! Any changes after they are provisioned exist only as changes (deltas) on disk.
It can save you on:
You should back-up everything that is imperative to your business.
-
Programs
-
Servers
-
Databases
-
Data
-
etc.
Short Answer
To summarize, there is a 56% (for a 1TB drive and potentially larger) chance of not being able to read every single sector or recover your single parity RAID with large SATA disks. That is why we recommend RAID-DP, SAS or FC. If you require SATA or ATA work at
Long Answer
“As disk drives have gotten larger, their reliability has not improved, and, more importantly, the bit error likelihood per drive has increased proportionally with the larger media. These three factors—larger disks, unimproved reliability, and increased bit errors with larger media—all have serious consequences for the ability of single-parity RAID to protect data.” (IMPLEMENTATION OF RAID DOUBLE PARITY FOR DATA PROTECTION TR-3298 [12/2006] , Chris Lueth, Network Appliance, Inc. )
“With Seagate, every SCSI/FC drive has a bit error rate of 1 in 10^16, except for the Cheetah 10K.7, which is at 10^15 but maxes out at 300 GB. With Hitachi, every SCSI/FC drive has a bit error rate of 1 in 10^16. With Seagate, Hitachi and Western Digital, every SATA drive has a bit error rate of 1 in 10^15 or even 10^14, even with terabyte capacity drives! Not one exception.
This means that if (and only if) you’re using [single parity] RAID, SCSI drives are hands-down more reliable. SCSI drives have a better bit error rate, and are smaller in capacity (so less data needs to be read for a RAID rebuild). In the worst case, the Cheetah 10K.7 drives, there’s only a 1.7% chance that a block will be lost in reconstruction. In the SATA worst case, the Hitachi Deskstar 7K1000.B or the Seagate DiamondMax 22, there’s a massive 56% chance the RAID rebuild will fail.” (Are Fibre Channel and SCSI Drives More Reliable?, Jered Floyd , August 20 2008, http://permabit.wordpress.com/2008/08/20/are-fibre-channel-and-scsi-drives-more-reliable/)
Raid stands for Redundant Array of Independent Disks.
For a good introduction on RAID see http://en.wikipedia.org/wiki/RAID
Depending on the type of RAID used it can be GOOD or BAD.
No Parity RAID
RAID-0
RAID-0 is a non redundant array. It keeps a single set of data across all of your disks. It is used especially in performance applications where storage is expensive and the risk of losing your data is of little or no concern (a single failure will corrupt/destroy the array).
RAID-1
RAID-1 is a mirrored array (requires an even number of disks). Your data set keeps a mirrored copy (you are losing 50% of your disk space) of itself on every second drive. It is costly, although provides performance benefits (in operating systems supporting split seeks). RAID-1 is vulnerable to a double disk failure.
RAID-10 (RAID-1+0)
RAID-10 is a combination of RAID-0 and RAID-1. RAID 10 provides better rebuild performance and slightly better fault tolerance (for more than a single disk failure) than RAID 01. RAID-10 is still susceptible to double disk failures (failure of mirrored set and vulnerable to BER ) . RAID-10 is recommended (depending on use).
For a more detailed description on nested RAID levels read: http://en.wikipedia.org/wiki/Nested_RAID_levels.
Single Parity Raid
In business critical storage, single parity RAID is not generally recommended. It is vulnerable to double disk failure, which is significant (see Which disk should I use?) and the performance degradation during a rebuild is extremely poor. Any additional failure during a rebuild of a “degraded” (the RAID is rebuilding due to a single failure) RAID set will probably cause complete data loss. The larger the drive the larger Mean Time To Repair (MTTR). The longer this takes the more you are at risk of losing all your data.
It is the cheapest redundant RAID solution (per megabyte).
Parity is used for redundancy instead of mirroring data. Parity is not data – it is a calculation – which can be used to recover data. It allows for less unused and redundant storage. Parity or RAID needs to be re-calculated in the event of a single failure:
-
which causes a major performance degradation
-
writes to disk require parity calculations which can seriously slow your system
-
recovery can take days
-
as the calculation is derived from all the disks during a recovery, it is extremely vulnerable to the BER.(...read more) which could lead to the likelihood of corruption and disaster.
RAID 4 & RAID 3 & RAID 2
RAID 2 (bit oriented), RAID 3 (byte oriented) and RAID 4 (block oriented) stripe data across disks and typically have an additional parity drive for redundancy. It suffers all of the vulnerabilities as mentioned above.
Raid 5
RAID 5 distributes the parity over all of the disks. This gives performance benefits for normal operation (reads) . It suffers all of the vulnerabilities as mentioned above.
Double Parity Raid
Double parity RAID systems allow for double disk failures. More failures will result in data loss. Double parity RAID systems typically rely on features of single parity RAID systems.
RAID-6
RAID-6 allows double disk failures by using additional parity information. Typically the reads are similar but the write performance is worse than RAID-5 by more than 30% as it requires a second calculation (http://web.archive.org/web/20060511183621/http://www.virtual.com/whitepapers/HDS_Using_Raid-6.pdf).
RAID-DP – The best storage method available
RAID-DP (media.netapp.com/documents/netapp-raid-dp.pdf) is a proprietary high performance implementation of RAID-6 written by NetApp (http://netapp.com). RAID-DP requires two dedicated parity disks in every RAID group (one for parity, the other called diagonal parity). It is more resilient than RAID-10. It is faster than RAID-1 and is as fast if not faster than RAID-10. It doesn't suffer the write performance constraints of typical RAID-6 implementations due to the in memory allocation of chunked data and calculation of parity before data persistence onto disk. A non volatile file system allocation data-chunk (tetris) is processed by the file system (using the Write Anywhere File Layout (WAFL)) and is passed to the software RAID layer (which performs the parity calculation). After the parity calculation is made, the information is then persisted on disk (with parity) and atomic checksums ensure consistency of the written data. If any failure occurs during this process, the system state can be recovered (from non-volatile in memory logs (NVRAM) even on another machine!). As WAFL always writes blocks to new memory locations (instead of replacing old ones) it allows random writes to work sequentially. RAID-DP also interestingly allows disk firmware updates in real-time without any outage. http://media.netapp.com/documents/tr-3574.pdf.
“The performance of RAID-DP volumes is comparable to that of RAID 4. Read operation performance is exactly the same for each type of RAID group. Depending on the type of write, performance on RAID-DP can be about 2% to 3% slower than that of RAID 4. The reason for this small performance difference is that an extra write occurs to the second diagonal parity disk on RAID-DP volumes. There is no discernible impact to the CPU utilization from running RAID-DP versus RAID 4.” (For more information about RAID-DP, see “Implementation of Raid Double Parity for Data Protection”, Chris Lueth, Network Appliance, Inc., http://www.netapp.com/library/tr/3298.pdf)
RAID-DP.
We will never recommend disk backups for a business unless you just can't afford anything else.
-
A USB disk backup is equivalent of offline version of RAID-1 (a single parity RAID). Depending on your disks, it could be likely (8%) that you never recover your data.
-
backups are hard to test, you may be copying onto an unreliable (corrupt drive)
-
they are difficult to manage
-
they are difficult or impossible to effectively version control
-
they won't help in the majority of business continuity requirements, time to recovery is prohibitive
-
what if your drive dies while backing up?
-
people don't in general do differential backups so are wasting resources
What about backing up with optical media?
As heat and age can effect the longevity of optical media, we don't recommend it for serious or long term backup of business critical data. Their lifetime can be as short as 2 years (http://www.itl.nist.gov/iad/894.05/loc/).
Cheap online services are not typically suitable for business use and nusiness continuity as:
-
expect to lose your data. The cheapest systems operate with the cheapest model, which is a cocktail for disaster and corrupt data. See Single Parity Raid for an explanation.
-
RAID-4 or RAID-5
-
ATA (SATA) disks
-
large disks
-
not all files are necessarily backed up with cheap online tools
-
you can never be 100% sure your data wont be compromised and there is also the possibility of the company disappearing we can provide you with your own hardware off-site to ensure complete independence and autonomy
-
a local repository where you can (if required in an emergency) get to your data and do a massive restore (>1TB)
-
a local repository where you can download at optical fibre speeds
-
work with your data online offering complete application and data fail-over
-
not all cheap alternatives keep versioning history, so if you overwrite a file with a corrupted one you could lose everything
-
files aren't backed up differentially (they take up a lot of bandwidth) and take up a lot of space
Yes – even if you were in the jungle using a satellite-phone. Although depends on the amount of data you need to transfer.
Zero downtime or high availability is a term to describe solutions which never go down. It is possible to keep your I.T. systems with “near perfect” uptime. Ask us how.
This depends on your requirements and business impact analysis. Contact us.
This should be based on your requirements. Ask us for assistance.
No? You need a business impact analysis. Ask our business continuity consultant for more information.
Yes? It is likely we can help. Ask us for some assistance to see whether it is possible.
For an example if you had 80 Windows file servers, on the implementation we could provide you, it is expected that you'd get 139% Return on Investment (ROI) after 3 years. Which is almost $1.5 Million in cost savings!
Saved in:
Other cost savings can be gained from:
-
unifying existing SANs
-
reducing downtime
-
etc.