Disaster Recovery and Fibre Channel Driven Data Center Technology Innovation
Recently, Henry Newman, Chief Technology Officer of Instrumental and author of the Enterprise Storage Forum, pointed out that many large organizations have a strategic planning process, trying to clarify which technology they need in the next three to five years and set a long-term budget for this process. Of course, the problem is that complex businesses do need to predict where the technology will develop in the next few years and there will definitely be new technologies hindering these plans in the next few years. Early detection of disruptive or innovative technologies, or at least finding out where these technologies may appear, will at least reduce unexpected situations. This planning process allowed me to consider why the technology has evolved and what unexpected things will happen to the storage network market in the next few years.
"Disruptive technology" is a clearly overused vocabulary. There are very few truly disruptive enterprise technologies, but this does not mean that there are no important technological innovations. Let us look at two examples-tape storage and Fibre Channel. In these two areas, technical problems have driven the creation of solutions, and then the solutions have significantly changed the enterprise data storage environment.
Tape and Deduplication In the past 10 years, no market has changed more than the data backup and recovery market. Just 10 years ago, backup only included backup directly from hard disk to tape. Like the network, the tape speed is very slow. Within two years, the tape speed increased to 35MB per second locally, and the compression speed exceeded the network performance. Two tape drives can saturate the network, especially when compression is used. There is an imbalance between storage and network. Considering the delay of the tape, when there is no multi-speed tape drive, the data must be transferred to the tape in order to get better performance. Since the network speed is slower than the tape speed, three innovative technologies have been generated due to customer complaints:
1. Hard disk to hard disk to tape backup.
2. Tape library manufacturers have developed virtual tape libraries.
3. Tape drive manufacturers have developed multi-speed tape drives.
These innovative technologies address the impact of slow network performance on backup and recovery. Even backup / restore vendors in the hardware and software industries have developed this technology to resolve user complaints. You can assume that this problem is solved. However, network performance has not improved. By 2005, many customers could not complete the backup within the prescribed time window. This has promoted the application and development of deduplication. Deduplication reduces the amount of data that needs to be backed up and stored. Some people may call deduplication a disruptive technique. However, I think this is just a very innovative technology, which is the logical next step that occurs after the customer fails to complete the backup in time. Deduplication may be innovative, but it is not much different from standard file compression methods unless you perform cross-file compression instead of compressing in one file. In fact, when deduplicating pioneer data
When Domain (now acquired by EMC) appeared, it called its technology "global compression." It called the technology deduplication only three years later.
If the IT environment can complete its own backup in the required time period, people may not open up the technology of deduplication. If you can provide 10 Gigabit Ethernet at a reasonable price in 2005, you may not see a significant investment in deduplicating hardware and software, because streaming tape will work just as well. However, deduplication has another result: by reducing the cost of the hard drive, it is close to the cost of tape, thereby making the role of tape more and more perform archiving tasks. This achieved results that it did not anticipate. If tape sales continue to decline, what will happen to the backup market that still requires tape, what will happen to the huge archive market that requires tape, and where most data (if not all) cannot be deduplicated? Do n’t think This is a small business problem, and some of the largest institutions in the world are users who use tape archives heavily. One problem is that when a technology market slows down, as a strategic planner, figuring out what will fill this market gap will become a guessing part of the planning process.
If manufacturers fill market gaps as required, what is the next step in tape? The slow development of the network opens the door to hard drives and deduplication in the backup market. However, network standards organizations are now working overtime. The application of 10 Gigabit Ethernet is accelerating. The 40GB Ethernet and 100GB Ethernet standards are not far off. It is too late to reserve tape for the backup market, and technologies such as deduplication and hard drives have inevitably appeared. This means that the tape market will be under constant pressure and some consolidation may occur, because a considerable portion of the tape market is still used for backup and recovery. Over the past few years, the proportion of tape shipments used for archiving in this market has increased significantly. However, although archiving requirements have grown, this growth is not enough to compensate for the decline in tape shipments in the backup market.
This will have a profound impact on the archive market. Without a large amount of profit, it is very difficult to carry out technological innovation. As the market shrinks, profit has always been a problem. Maybe in the next few years this market will shrink to only a few manufacturers, or maybe slow down the pace of technological innovation. It would be nice if holographic storage could come out to solve this problem, but we have been waiting for it for decades.
Fibre Channel meets 10 Gigabit Ethernet Fibre Channel product development is parallel to tape, because both technologies rely heavily on the pace of Ethernet technology innovation. The difference is that Fibre Channel has benefited from the lag in Ethernet performance, but this situation may change.
Fibre Channel has been around since 1996. At the time, I was the first to use this technique. At that time, the speed of the Fibre Channel was 1GB, not based on the structure, but based on the circular type. When Gigabit Ethernet rises at a reasonable price, 2GB Fibre Channel will also enter the market. Fibre Channel has happily evolved from 2GB to 4GB, but Fibre Channel technology has been ousted for two reasons: SATA is rapidly growing as an enterprise technology; the Ethernet community has decided that it needs to make huge changes.
The SATA interface for hard drives is not very reliable, there is a higher error recovery delay and a lot of things that affect reliability and performance. However, the SATA interface is still used and the price is now cheaper. Hard disk manufacturers use Fibre Channel as an enterprise technology, but this cost is too high and the interface is not compatible with SATA interface. This result is a combination of SAS and SATA interfaces (using a connector) and hard disk manufacturers can use a chipset for SAS and SATA interfaces. This significantly reduces costs. We keep hearing about 16GB Fibre Channel, but I doubt it can gain a lot of market share because FCoE (Fibre Channel over Ethernet) on 10 Gigabit Ethernet will become a more attractive option. Fibre Channel can be used for a long time in part because of the slow development of Ethernet. However, this situation is finally about to change. Tape and Fibre Channel may have been affected differently by the long-term lag of Ethernet performance. However, their destiny may be the same.
IT technology innovation comes from demand and commercialization. The lack of technology innovation can promote the development of the technology market as well as technological innovation (Gigabit Ethernet lasts too long to open the door for hard disk backup and deduplication). The commercialization of technology is another long-term trend that has created certain technological fragility. What this matter means to you depends on your technical planning window. I did not see all the changes due to the long residence time of Gigabit Ethernet. However, I do acknowledge the limitations of Fibre Channel. Although Fibre Channel was vigorously promoted in early 2000, it was not applied to the motherboard of computers. Once this happens, it is obvious that Fibre Channel will one day become a secondary thing. The only surprise is how long it takes Ethernet personnel to make this happen.
Trying to predict the IT market in your strategic planning process is an art, not a science. The opinions of analysts usually reflect the views of the vendors they are close to. The best tools to understand and plan for the future are your eyes and ears, or if you need to do long-term strategic planning, at least find an independent help.
Hdmi C Connector,Mini Hdmi Connector C Type 19Pin,C Type Copper Connector,Usb3.1 Type-C Converter Hdmi
Dongguan City Yuanyue Electronics Co.Ltd , https://www.yyeconn.com