How to classify “cloud garbage”?
Editor’s note: This article is from WeChat public account “Big Data Digest” (ID: BigDataDigest), author: Cai Jie, Linan An, Cao Peixin.
On July 1 this year, the “Regulations on the Management of Domestic Waste in Shanghai” was officially implemented. It was called “the most stringent garbage classification measure in history”.
As soon as the regulations came out, the barrels of these four colors completely changed the lives of the people of Shanghai.
But a new kind of garbage – “cloud garbage”, may not have been heard by everyone.
In fact, “cloud garbage” is also a term that has just appeared in recent years. It refers to the various old IT infrastructures that have been eliminated due to equipment upgrades in data centers that provide cloud computing services.
Maybe this kind of garbage ordinary family will not be produced, and it is far from people’s lives, but don’t underestimate this part of the garbage, this is a huge market worth 300 billion US dollars!
But it is not easy to recycle this kind of garbage. Because it involves many hard disks with user personal data, improper handling is very likely to cause data leakage! Many giant technology companies are guilty of this, and even spawned a third-party company that specifically recycles this “cloud garbage.”
The global scrap metal market is turning its attention to cloud computing equipment recycling
As we all know, we are now in an era of data explosions, the global demand for data storage and cloud computing is growing, and the number of large data centers is also exploding. Citigroup’s research shows that since 2015, cloud infrastructure investment has grown substantially, and the market annual growth rate of data center related equipment is expected to increase by an average of 16% this year and next.
However, these data centers are both a place for massive data processing and a giant garbage dump.
Why do you say that? A report in the enterprise network lists such a set of data. A medium-sized data center generally has about 3000 racks, and a rack can hold 10 servers.
This medium-sized data center can hold about 30,000 units of equipment, with a building area of 2,000 square meters. These devices have a service life of five years. Nowadays, data center technology is developing very fast. Many devices have been eliminated early in less than five years. The equipment that has been eliminated has become waste and has become electronic waste.
This means that the earliest batches of equipment have exceeded their useful life and become a huge amount of e-waste. And the current state of high-speed iteration of this technology is creating a new market – cloud recycling. It is estimated that this market is about $300 billion.
The large US recyclers Sims Metal Management Ltd. (Sims) and Electronic recycling International Inc. (ERI) are among the beneficiaries.
Some devices in the data center (such as processors, fans, etc.) can be removed separately and resold on the electronics market. They also contain metals such as aluminum, copper and steel, and these metals become more valuable as large economies such as China rely more on consumption (rather than exports) to drive economic growth.
Sims and ERI are involved in the US to handle a small amount of cloud computing material recovery pilots, and predict that the amount of recycling will increase significantly in the next few years. Rising trade barriers and weak commodity prices have hit corporate profits, and cloud recycling may be the new profit growth point for these recyclers.
At Sims’ factory in West Chicago, Ill., staff are detecting and renovating laptops. Image source: “Wall Street Journal”
It’s not just a simple sale.
2018, AsiaThe annual capital expenditures of Mason, Microsoft and Google totaled more than $60 billion, a 50% increase from the previous year. All three companies are running ultra-large-scale cloud computing networks and data centers, and are constantly upgrading and expanding.
Technology giants stumped by “cloud garbage”
With the concern about the security of sensitive data, technology giants such as Amazon, Microsoft and Google are cautious about the disposal of discarded equipment.
Some companies require recyclers to disassemble servers on site and film the entire process; others are hiring security guards with gun protection on the way to a waste server. Coupled with the industry’s differing views on the definition of “recycling materials”, some technology companies even require that all equipment be thoroughly comminuted and melted, even if it contains reusable parts.
Google compares its data center to a “small city” full of servers, drives, routers and other components. Google is eager to recycle and resell parts as soon as possible to avoid sending them to the dump. It may be perceived by the public that this will reveal user data, and both Google and Amazon declined to comment on plans to update the infrastructure.
In the Sims factory, workers are checking electronic equipment. Image source: The Wall Street Journal
Microsoft has also made great efforts to recycle, refurbish and resell cloud infrastructure components, trying to make the process more sustainable and safer.
Alistair Field, CEO of Australian listed company Sims, believes that data center recycling transactions will occur within 12 to 18 months and will gradually improve over the next few years.
He estimates that about 20 million tons of waste equipment will be dismantled and recycled every year from 2025. And for professional third-party recycling companies to do, Field said: “The key to the pilot is how to manage the entire process, such as how to ensure that every step of the recycling process is safe? Which processes customers want to participate in? “SSD hard drive can be found even if it is physically destroyed
If the company decides to physically destroy the disk, then the destruction of the SSD drive will cause more trouble.
Fredrik F, Vice President, Corporate and Cloud Elimination Solutions, BlanccoOrslund pointed out: “Even if the SSD drive is destroyed, data may still be found.”
Although very large-scale cloud computing providers like Microsoft Azure and Google Cloud use shredders to process drives (Google uses robotic automated destruction), this is rare in traditional enterprise data centers and is quite expensive.
The robot destroys the hard drive in Google Data Center. Google Data Center Vice President Kava is shown in a demo of Google Cloud Next 2018. Image source: Google
And doing this requires a full audit trail, which means you need to take the serial number of each drive and the shredding process. In addition, there are some companies worried about security trying to use a magnetic degaussing device to erase data, rather than physically damaging the hard drive, but today’s hard drive shell will block the magnetic, SSD hard drive data will not be affected by the degaussing device Impact.
Although hiring third-party services can provide effective physical damage through tracking audits, this increases the cost of protecting drives during transmission. “Users need service providers to safely ship hard drives to their facilities,” Forslund said. “Many security teams will not allow such transportation. If allowed, there are strict requirements, such as transportation by armed escort vehicles, which requires a lot. Money, which is why high-volume hard drives are expensive.”
Deleting a key is not a foolproof
More and more companies are trying to eliminate the password of the hard disk by simply deleting the key, because this solution is fast and simple. According to the study, 64% of companies adopt this approach, and this proportion is even higher in regulated industries such as medical and pharmaceutical.
However, Forslund warned: “This is not a foolproof measure. Users need to encrypt the data of the storage device all the time, because if it is open at any time, data leakage may occur. Users need a more complete system to manage encryption. Key.”
If this is the case, it is important that the verification and audit key has been deleted, even if the hard drive’s data has been deleted and the management tool marks it as unopenable. But only 70% to 80% of the time, the old hard drive can be deleted to an auditable level, which means the departmentThe data can still be recovered from it with the right tools.
is both a huge market and a huge challenge. The recycling of “cloud garbage” still needs to be cautious
In 2017, a report jointly released by the United Nations University, the International Telecommunication Union and the International Solid Waste Association stated that in 2016, only 20% of the world’s electronics The garbage was properly recycled.
The report points out that although current cloud computing technologies can improve the efficiency of device use, such as one device can provide more services than ever before, the increasing demand for cloud computing means that people need more data centers. It also produces more e-waste.
The ERI recycling facility in Indiana and Massachusetts is currently using manual cleaning of e-waste. John Shegerian, co-founder and executive chairman of ERI, believes that the number of e-waste will continue to increase and he hopes to use robots in the future. In the field of electronic waste recycling, ERI is also actively seeking new partners.
Every day, ERI’s Holliston factory smashes about 33 tons of cathode ray tube TVs, 6 tons of flat-panel TVs, and 7 tons of computers. Image source: Made
But the biggest challenge for huge profits is whether the data will leak. In the past, e-waste recycling was more of a personal device. Since it was decided to sell, individuals should be responsible for their actions. But now a company that stores a lot of user data has to consider the risk of data leakage if it considers selling these devices.
Slightly inadvertently, it would violate the European New Data Act GDPR or the California Consumer Privacy Act CCPA.
China has introduced laws and regulations such as the “Solid Waste Pollution Prevention and Control Law” and the “Clean Production Promotion Law” on e-waste, but it is more concerned with the problem of e-waste polluting the environment, but it does not involve it. The data.
As the data center of domestic technology giants is updated more frequently, how to ensure that data is fully protected while the resulting “cloud garbage” is being recycled.