Ah, the wide and deep and far-fetched wonders of storage – IBM researchers talking about chemically driven phase transformations, Backblaze looking at a Pelican, HPE setting up an Azure Stack and SQream going off to Thailand.
New research, updated products, new deals, new hires and new money are all in this week's worth of storage goodies. Enjoy.
Backblaze and Pelican
Gleb Budman, co-founder and CEO of Backblaze, gave us his thoughts on Microsoft's Pelican archival disk storage.
El Reg How do you view the use of high 52U rather than standard 42U racks and the weight issue (3,000 lbs/rack)?
Gleb Budman Fine if your data centre can support the weight (e.g. on a concrete slab w/o raised floors.)
Most can't, so limits your options for not a lot of benefit.
El Reg What is your view on the use of front/bottom to rear/top cooling?
Gleb Budman Perfectly reasonable. Not significantly different than what's common though. Data centres with raised floors have cold air coming from the bottom and sucked out the rear.
Might be minutely more efficient to have the heat exit on top.
El Reg How about the use of MAID Technology to lower power-draw to 3.5kW/rack. It seems to us you get a latency roughly halfway between disk and tape. Is this what users want?
Gleb Budman Depends on the use case. Doesn't work for ours. We have evaluated power-cycling and putting drives to sleep in the past.
However, between users wanting to access certain files and the system doing housekeeping, it hasn't made sense.
For our cloud storage service, B2, we also wouldn't want to limit the fundamental access to data since it's a real-time service.
Having said that, some users may have use cases to fit this pattern. I doubt most cloud-scale companies would, but enterprise IT folks may.
El Reg Pelican is evaluating shingled and non-shingled disks. What are your views on using shingled and non-shingled disks?
Gleb Budman SMRs have really bad rewrite performance, and would require some pretty major software/hardware tweaks to use effectively. I'm sure Microsoft has found that out in their evaluations. The Linux kernel is starting to add workarounds for some of this with smart filesystem bits, but they're still very early on.
The SMR disks are priced lower, so that's compelling, and if the use cases are literally replacing tape (i.e. write once, rewrite never, read almost never) that's probably a reasonable choice.
El Reg Pelican is using object storage (talks of storing blobs and using Gets and Puts). Does Backblaze use object storage? What are the arguments for and against the storage technology you use?
Gleb Budman Yes, the Backblaze cloud storage is an object store. Our cloud backup service uses that underlying; our B2 cloud service exposes that via APIs.
In general object storage is designed for scale in a way that file and block storage systems are not (we currently store billions of objects and are on-track to have an exabyte of storage this year). Object storage also supports having metadata for objects. However, object storage is generally not as fast as block storage.
El Reg Is there a real need (that Backblaze sees) for a storage technology that is faster than tape and slower than disk, but nearer tape cost than disk cost? If there is then would Pelican meet that need?
Gleb Budman As you know, Backblaze believes strongly in driving down the cost of storage, and that lower-cost storage enables new use cases and businesses. Technologies that help drive down costs help us all. However, storing data that you won't be able to use – either because it's too slow or too expensive to get at – is much less useful.
We've looked multiple times over the years at storing data on tape, but in a cloud environment where customers want to be able to access the data, it never penciled out. Replicating a model where it's difficult to get the data with hard drives isn't necessarily a solution. There are use cases (primarily archiving) for such a technology if it really is very inexpensive. However, without numbers from Microsoft, it's hard to say whether that is true.
There are two costs:
- CapEx: Ask Microsoft what the capex for Pelican is. (About halfway through this post (Storage Pod 6.0), you can see the $/GB in capex for each model of hardware we had at the time; obviously lower now.)
- Operating Costs: Data centre space and power account for about a third the cost of the capex in a cloud storage system. So, if capex goes up by 50 per cent, you'd have to get the data centre space and power for free to break even.
If Pelican costs less in capex, and less in operating costs, the archiving use case could be a reasonable one. If not, it's going to be hard to argue for it.
Nature Electronics, 17 April, ran an IBM Research paper, Mixed-Precision In-Memory Computing, which talked about using Phase-Change Memory (PCM) in a compute-in-memory application.
The abstract says:
We introduce the concept of mixed-precision in-memory computing, which combines a von Neumann machine with a computational memory unit. In this hybrid system, the computational memory unit performs the bulk of a computational task, while the von Neumann machine implements a backward method to iteratively improve the accuracy of the solution.
The system therefore benefits from both the high precision of digital computing and the energy/areal efficiency of in-memory computing. We experimentally demonstrate the efficacy of the approach by accurately solving systems of linear equations, in particular, a system of 5,000 equations using 998,752 phase-change memory devices.
The article text states: "These [PCM] devices form the basis of in-memory computing: an approach in which both information processing and storing computational data are performed on the 10 same physical devices organized in a computational memory unit.
"With such systems, various physical mechanisms, including Ohm's law and Kirchhoff's circuit laws, chemically-driven phase transformations, the pattern dynamics of ferroelectric domain switching, and the physics of crystallisation and melting in phase-change materials, can be used to perform arithmetic and logical operations."
And here were we thinking a chemically driven phase transformation involved ingesting illegal substances.
We leave you to enjoy the paper while we move on to the more understood topic of backup with Nakivo.
Nakivo Backup and Replication v7.4 adds automated VM failover using replicated VMs in a DR site and a near-instant automated failover job.
It can recover accidentally deleted or corrupted files directly to their source VMs or a different location. The fast recovery is performed from deduplicated VM backups, without recovering the entire VM first.
The new software can be used to protect AWS EC2 instances and store the backups on site or in the cloud. The number of recovery points per EC2 instance is 1,000, meaning greater reliability and more recovery options.
During business hours, network admins may want to limit the bandwidth used by data protection processes, and this v7.4 software allows limits to be set on a per-job basis, ensuring enough bandwidth is available for other business applications.
If something happens to the VM or physical server hosting the VM backup software, a new instance of Nakivo Backup and Replication can be installed in under a minute. As manually reconfiguring all of the backup settings might be time-consuming, the software automatically backs up the entire scope of settings made in the product's web interface and saves such self-backups in available backup repositories.
Once a new instance is installed, Backup and Replication can import all the previously established settings from the backup repository, including jobs, inventory, schedules, and preferences.
V7.4 introduces a Global Search feature for finding any item, such as a job, a repository, or a transporter. There is also the option to instantly perform individual or bulk actions on such items directly from the search results page.
Flash VM Boot can near-instantly boot virtual machines directly from deduplicated and compressed VM backups. Such VMs can then be used for sandbox testing or other purposes. V7.4 extends this to Hyper-V VMs.
After each backup or replication job, the product can automatically test-recover the VM and take a screenshot of the operating system, which is attached to a report and sent via email. This Screenshot Verification feature is now available for Hyper-V as well as VMware VMs.
The product can truncate Microsoft SQL Server transaction logs on the source VM. These logs are truncated automatically after each successful VM backup or replication job, preventing them from taking all available disk space and causing a server crash.
It has instant recovery of Microsoft SQL Server objects, such as tables and databases. The granular recovery is performed from deduplicated backups, without requiring the full VM to be recovered first, which saves much time.
The web interface of Backup and Replication v7.4 includes a built-in live chat facility with Nakivio's Technical Support team.
HPE has a ProLiant for Microsoft Azure Stack (Gen 10) server system, based on a DL380 base. It has up to 120TB raw capacity per node, Silicon Root of Trust, integrated HPE OneView, and pay-as-you-consume pricing with HPE GreenLake Flex Capacity.
Hyperconverged system vendor Pivot3 enjoyed a more than 60 per cent increase in bookings in Q1 2018 compared to Q1 2017. Large enterprise and Fortune 1000 customers represented 67 per cent of its Q1 sales. It saw a 56 per cent increase in customers using its HCI to support multiple mixed-application workloads from Q4 2017 to Q1 2018.
Samsung has begun mass producing 10-nanometer (nm)-class 16Gbit LPDDR4X DRAM for automobiles; 10nm-class being a process node between 10 and 19 nanometers.
Scality and Wasabi announced the integration of Wasabi hot cloud storage into Scality's Zenko multi-cloud data controller, along with a new marketing and technology partnership. Wasabi is the newest back-end cloud storage service supported in Zenko.
ScyllaDB, a real-time big data database startup, has closed a $10m Series C funding round led by TLV Partners. Previous investors Bessemer Venture Partners, Qualcomm Ventures, Magma Venture Partners, and Western Digital also participated in the round. This latest round raises ScyllaDB's total funding to $35m.
Seagate has achieved Common Criteria (CC) certification across a portfolio of its disk and solid stare drives, expanding on its existing Seagate Secure Certified Service Level. The Seagate Secure Products, including the Nytro, Exos, and Barracuda portfolio, along with this certification ensure all solutions meet stringent data privacy laws and regulations, including HIPAA and GDPR, set to be enforced on 25 May.
Sphere3D is holding a special shareholders meeting on 31 May where shareholders will vote on the sell-off of Overland Storage to an as-yet-unnamed owned by Sphere3D’s chairman and CEO, Eric Kelly. The rump Sphere3D gets renamed to HVE ConneXions, Inc. and Kelly leaves to run Overland Storage,
Storage Made Easy has an interesting blog post: "If data is the new oil, then metadata is the refiner."
Zadara, a provider of enterprise-class storage-as-a-service (STaaS), says its Zadara Storage Cloud and VPSA services meet the European Union’s new General Data Protection Regulation (GDPR).
DR seller Zerto has a cross-selling agreement with Microsoft, meaning Zerto sales teams and channel partners are incentivized to co-sell and cross-sell Zerto's IT Resilience Platform and Microsoft Azure, aligning its enterprise sales motion with Microsoft sales teams.
BlueData (Big Data as a Service) has appointed Shane Margraves as VP of worldwide sales. He comes to BlueData from Dell EMC.
Datrium has appointed Phillip Liu as VP of Engineering. Previous incumbent Jana van Greunen leaves to a new post at Facebook with Datrium's blessing: "We wish Jana the best at Facebook, and are super excited by Phil's strong experience in building products on public clouds – the foundation for our newer initiatives building on Cloud DVX."
Composable server startup DriveScale announced the appointment of veteran technology executive from NetApp and Pure Storage, Brian Pawlowski as its Chief Technology Officer.
Druva has hired Sherry Lowe as its chief marketing officer. Previous CMO Matt Morgan had gone upstairs to advise the CEO and board. Previous VP corporate marketing Wendy Perilli had left to join Github.
ESG analyst Jason Buffington has joined Veeam as Senior Directir for Product Strategy.
SQream, developer of SQream DB, a GPU-accelerated database, has entered a partnership with Thailand's top mobile operator AIS. It follows deals with Alibaba Cloud and India's ACL Mobile. ®