Update: May 26, 2020. Customer availability note added.
Dell Technologies has added the scale-out Isilon OneFS file server to Google Cloud. Customers get 50PB file name space and 200GB per second aggregate throughput, which is managed by Dell Enterprise Services and billed through Google.
OneFS for Google Cloud provides terabit networking and sub-millisecond latency between the Isilon hardware and Google Compute Engine services. The annual subscription service can count toward any committed Google Cloud spend.
The OneFS software has not actually been ported to run on Google hardware within Google Cloud. The planting of Isilon filers in to the Google Cloud is more akin to NetApp’s Azure NetApp Files, where NetApp filers are placed in Azure data centres.
Customers order the Isilon service via Google Cloud Marketplace. Customers are provided with Isilon Gen6 hardware and configuration and management is conducted through the Google Cloud console. Each Isilon cluster is dedicated to a single customer, so performance is predictable and data tenancy is assured. There are no egress charges associated with moving data between this storage platform and other Google Cloud services.
Existing applications using Isilon OneFS file storage can run in Google Cloud and use OneFS in the same way.
A Dell Technologies white paper shows how OneFS for Google Cloud can be used for deep learning and illustrates an example setup:
Isilon hardware communicates across a Google Virtual Private Cloud (VPC) for access by Google compute instances. VPCs are global resources and are not associated with any particular region or zone.
Dell Technologies did not provide any availability information. Blocks and Files suspects there will be geographic availability restrictions.
Competition and Performance
Isilon’s OneFS joins Filestore, NetApp Cloud Volumes, Google’s Elastifile and Panzura as file server suppliers in GCP. Qumulo is also available for Google’s cloud but listed separately. In that listing Qumulo claims it’s the fastest and most affordable file system in the cloud. This claim may now be out of date.
ESG used the IOzone, an industry-standard benchmark, and Scratch DNA, a benchmarking tool that emulates the write-intensive storage workloads typically generated in genomics research. It tested a 2PB OneFS file system with peak 200GB/sec aggregate read throughput and 121GB/sec aggregate write throughput on IOzone (1024 threads). Scalability was linear. The peak Scratch DNA aggregate write throughput was 110 GB/sec.
Response time was under a millisecond. ESG compared this to a mystery Vendor X competitive NAS supplier running on Google Cloud. It refused to identify who it was, and acknowledged that vendor’s release numbers used a smaller workload and smaller filesystem. That said, OneFS offers up to 500x higher file system capacity, 46x maximum read throughput and 96x higher maximum write throughput.
A 500X higher capacity, based on OneFS’s 50PB limit, would mean a 100TB limit. NetApp Cloud Volumes offers up to 100TB. Qumulo on Google Cloud scales to the petabytes level. Filestore scales up to 63.9TB. We couldn’t find a capacity limit for Panzura but nonetheless we think ‘Vendor X’ is NetApp and its Cloud Volumes product.
A NetApp Azure benchmark webpage talks of Azure NetApp Files delivering 30MiB of bandwidth with a 5TiB volume. It delivers around 1,600MIB/sec writes and 4,500MIB/sec reads on a Linux system. MySQL latency was between 1.8 and 4.6ms.
There’s no obvious apples-to-apples comparison here but our impression is that OneFS for Google Cloud is faster than Azure NetApp Files.
Dell is the top scale-out file system vendor in IDC’s Worldwide Scale-Out File-Based Storage 2019 Vendor Assessment, ranking above NetApp, IBM and Qumulo.
Update. A Dell spokesperson tells us that customers in North America, Singapore and Sydney will be able to take advantage of OneFS for Google Cloud at launch, with additional global locations to be announced based on customer demand.
Interview A set of containers that would need 20 servers to run in a VM-based environment would need fewer than 4 servers to run in a bare-metal Diamanti environment; stats like this make Blocks & Files want to know how and why Diamanti’s bare metal hyperconverged systems can run Kubernetes workloads faster than either single servers or HCI systems – so we asked.
Brian Waldon, Diamanti’s VP of product, provided the answers to our questions
Blocks & Files: How do containerized workloads run more efficiently in a Diamanti system than in a commodity server+storage system?
Brian Waldon: Diamanti’s Ultima Network and Storage offload cards and bare metal hyperconverged platform are the greatest drivers to the increased workload efficiency that Diamanti provides.
Diamanti D10 server diagram showing offload cards
Diamanti offloads network and storage traffic to dedicated embedded systems, freeing up the host compute resources to power workloads and improve performance. This minimizes iowait, giving customers a 10x to 30x performance improvement on I/O-intensive applications while getting consistent sub-100 microsecond latency and 1 million IOPS per node.
The use of offload cards in the data center is not entirely new – the AWS Nitro architecture has helped to deliver better consolidation for EC2 instances and produce more cost efficiency for AWS. However, by combining offload card technology with purpose-built Kubernetes CNI and CSI drivers and bare metal, optimized hardware (w/ NVMe and SR-IOV), Diamanti has an integrated software and hardware solution that removes extraneous translation layers that slow things down.
Blocks & Files: How much more efficient is a Diamanti cluster than a Dell EMC VxRail system on the one hand and a Nutanix AHV system on the other?
Brian Waldon: Generally, we are compared to virtualized infrastructure, and again here our offload technology provides far greater efficiencies. We have a holistic platform that is designed for containerized workloads. While we happen to use X86 servers behind the scenes we have eliminated a lot of the layers between the X86 server and the actual workload, which is another key component to the efficiency with our platform.
Given that we have a hyperconverged platform, Diamanti can use the bare minimum amount of software to achieve the same goals as competitive solutions and really drive efficiency. Our platform can drive up to 95 per cent server utilisation; alternatively, virtualised infrastructure can use up to 30 per cent of server resources just to perform infrastructural functions such as storage and networking using expensive CPU resources.
One of our customers was running an Elasticsearch application on 16 standard x86 servers supporting 2.5 billion events per day. They accomplished the same on just 2 nodes [servers] of the Diamanti platform while also reducing index latency by 80 per cent.
Another customer saw similar improvements with their MongoDB application – reducing their hardware requirements from a 15-node cluster to just 3 nodes of Diamanti while still getting 3x greater IOPS.
Separately, and with Redis Enterprise, Diamanti supported average throughput of more than 100,000 ops/sec at a sub-millisecond latency with only four shards running on a three-node cluster with two serving nodes.
Blocks & Files: How does the Diamanti system compare to industry-standard servers plus a storage array with a CSI plugin?
Brian Waldon: The key differences are the innovations behind the storage offload cards and the full-stack approach from CSI down to the hardware.
Because Diamanti has a holistic hyperconverged system we can provide guaranteed Quality of Service (QoS), access and throughput, from the perspective of storage, and honour that from end to end. We have a built-in storage platform to drive local access to disk while still providing enterprise-grade storage features such as mirroring, snapshots, backup/restore, and asynchronous replication.
Diamanti also minimises staff set up and support time by providing a turnkey Kubernetes cluster with everything built-in. We give customers an appliance that automatically clusters itself and provides a Kubernetes interface.
Comment
Diamanti claims a 3-node cluster can achieve more than 2.4 million IOPS and have sub-100μs cross-cluster latency. Basically it’s a lot more-bangs-for-your-bucks message. The company says it knocks performance spots off commodity servers running Linux, virtualised servers and hyperconverged systems running virtual machine hypervisors.
It’s software uses up less of a server’s resources and its offload cards get data to and from containers faster. Result; you run more containers and run them faster.
You can check out a Diamanti white paper about running containers on its bare metal to find out more. A second Diamanti white paper looks at Redis Enterprise testing.
Pivot3 is lowering cost per TB by running its video surveillance HCI software on a Lenovo gen 2 AMD EPYC server.
Cutting to the chase, the new Lenovo ThinkSystem SR655 server can support more video cameras, run more application software to process the images and stream data faster from the camera sites to a data centre than its Intel Xeon predecessor. That will help Pivot3 sell better against Intel-based HCI systems.
Customers also get significantly more bandwidth for video recording and 150 per cent increase in networking speed, with 25GbitE support for faster video transfer from the edge to the core.
The SR655 is a single socket AMD EPYC 7002, 2U server with up to 64 cores. This is claimed to provide equivalent performance to a dual-socket Intel-based server but costs less. For example, there is an opportunity for 50 per cent cost savings on software that is licensed per socket.
According to Pivot3, the EPYC processor is more powerful than the Intel Xeon used previously – a viewpoint it shares with HPE, which recently announced the SimpliVity VDI system,
The company claims the Pivot3/SR655 server delivers 10 per cent lower cost per terabyte, 14 per cent better capacity density, and supports PCIe 4.0, helping to deliver 100 per cent more bandwidth than similar servers using slower PCIe gen 3. The company is referring obliquely here to Xeon.
Users streaming data from AWS can save egress charges by storing and streaming the data using LucidLink’s Filespaces bundled with AWS S3 storage.
Peter Thompson, LucidLink CEO, has provided an announcement quote: “We … heard a great deal of frustration from our customers around the unpredictable nature of egress fees … we were able to come up with a bundle that will satisfy most customer requirements, assisting in cost predictability. As an added benefit, we eliminate the complexity of setting up a cloud service and pass on the combined buying power we create directly to our customers.”
The deal is that you buy a Filespaces subscription bundled with AWS S3 storage. Filespaces plus 1TB of S3 storage and 1TB of outbound (egress) traffic costs $85 per TB per month. Users are billed at $0.05/GB ($50 per terabyte) for egress costs above 1TB per month.
According to a Microsoft price comparison site, AWS S3 data egress costs $0.09/GB for the first 10TB with an initial 1GB free, and then decreases to $0.05/GB at 150TB or more. That makes the LucidLink 1TB+ egress charge of $0.05/GB look pretty good.
LucidLink claims it is making cloud services more accessible and predictable because users can now more accurately forecast egress fees.
The company is targeted cost-sensitive users who need good performance when processing files on-premises streamed from the cloud. Examples include video editors and others in the media and entertainment industry.
Veeam has joined forces with Kasten to offer cloud-native application backup to its customers. The deal spans sales, marketing and technology.
This partnership solves a problem for Veeam – how to get a foothold in cloud-native backup for cloud-native applications that run as a bunch of containers in a Kubernetes-orchestrated system.
Veeam aims to provide full spectrum backup to its customers, but has found that many are adopting a DevOps approach and running Kubernetes-orchestrated, cloud-native systems, sometimes with thousands of containers in clustered server systems. The company is facing a build, buy, or partner decision. Today’s outcome is its decision to partner with Kasten and its K10 product.
Dan Kogan, Veeam product and solutions marketing VP, writes in a company blog: “We’ve been seeing our customer’s interest increase in Kubernetes as a critical piece of their cloud infrastructure, creating a new area around container-native data protection that they need help in addressing. To that end, we’re excited to share that we are partnering with Kasten, the leaders in backup solutions for containerised workloads.”
Containers eat Veaam’s lunch
In the virtual server world, a virtual machine (VM) and an application are often effectively identical, and always closely intertwined. Veeam’s genius was to gain deep awareness of how hypervisors schedule and run applications in virtual machines and so provide better VM backup.
The company has made a killing in providing backup and restore under a data availability theme, first for apps in virtual servers, then virtual servers in the cloud and also physical servers. But containerisation and DevOps style development has encroached onto the virtual server world. Where does Veeam fit in?
A containerised application needs a backup facility that understands the backup focus is the application, not the container. Veeam could move with VM environments that migrated to the cloud – VMware Cloud Foundation, for example. But it had no capability for cloud-native backup for cloud-native apps.
Container backup
This has opened up opportunities for specialist startups such as Kasten, which recognised that Kubernetes and its orchestrating ilk provides the same framework for container backup as hypervisors do for VM backup.
In Kubernetes there is no mapping of applications to servers or VMs. Instead, applications are mapped to a set and sequence of containers. VM-focused or server-focused backup is blind in the Kubernetes world. Containerised app atomicity is the sine qua non here.
Bewildering containerised application environment
Comment
Veeam has signed reseller deals before, for instance with N2SW, to underpin Veeam Availability for AWS. This provided a cloud-native agentless backup and availability system to protect AWS apps and data.
It bought N2SW for $42.5m in January 2018. Problems ensued involving Veeam’s then Russian ownership, non-US domicile and N2SW’s federal business. Veeam sold N2SW in October 2019. Veeam is now US-domiciled, following the $5bn acquisition by Insight Partners in January 2020.
Kasten provides another route for Veeam into cloud-native backup. Will Veeam want to own its own containerised application backup technology? Blocks & Files thinks this is on the table, and Kasten, with total funding of $17m, is an inexpensive option.
It’s all about more and more speed and faster disaster recovery in this week’s roundup. Read on,
Nvidia DGX A100 GPU server
Nvidia has launched its third generation DGX GPU server, the DGX A100. This has encouraged third party suppliers to announce DGX support.
Iguazio, which supplies a high-speed, high-capacity storage platform for analytics-type workflows, including AI, was quick off the blocks. It says its data science platform technology supports GPU-as-a-service and has been certified DGX-ready by Nvidia.
Data scientists and engineers can work on a single self-service data science platform, which has horizontal scaling and dynamically allocates GPUs to workloads for greater resource efficiency. Iguazio’s open source projects, Nuclio and MLRun, serverless and orchestration frameworks respectively, manage GPU resources, and support stateful and distributed workloads.
Iguazio’s GPU-as-a-Service offering is part of the Nvidia DGX-Ready Software program.
HPC and general enterprise storage supplier DDN supports the DGX A100 with its A3I product. This originally paired DDN’s AI200 and AI400 all-flash arrays, running Lustre-based Exascaler software, with Nvidia’s DGX-1.
The A3I now works with NVIDIA’s DGX A100 and Mellanox switches and NICs. DDN says that A1400X-based A3I systems are deployed in Nvidia’s own data centre to provide high-performance and scalable shared file access for the DGX A100-based SuperPOD. This SuperPOD uses 140 DGX A100 systems.
DDN and Nvidia plan to release new reference architectures based on A3I and DGX A100.
Nvidia has said Dell, IBM, NetApp, Pure Storage and VAST Data storage lines will support the DGX A100.
InfiniteIO and Google Cloud Platform
InfiniteIO is now an official Google partner whereby its Hybrid Cloud Tiering software can be used to move infrequently accessed files from on-premises storage to GCP, saving on-premises file space and cost, but with no change to application access. The InfiniteIO system does all the file metadata processing usingFile Metadata Engine accelerator software and hides the file movement from applications.
InfiniteIO supports all four classes of Google Cloud Storage: standard, nearline, coldline, and archive. The upcoming v2.5 software release supports Native File Format for files tiered to object storage. Native File tiers whole files instead of storing the data in a proprietary format and allows customers to access tiered files directly via Google Cloud Storage.
The mdtest benchmark tests metadata performance of parallel filesystems using various simulated metadata workloads. InfiniteIO Blogger Kris Meier has described how an mdtest benchmark for a single-node InfiniteIO/Google Cloud Platform configuration reduced latency from about 1ms to 40μs.
InfiniteIO announced Hybrid Cloud Tiering functionality last month. It supports Amazon Web Services, Google Cloud Platform, IBM Cloud, Microsoft Azure and Oracle Cloud Platform. Currently supported on-premises NFS- and S3-based private cloud platforms include Cloudian, Dell EMC, HPE, NetApp, Scality and Western Digital.
InfiniteIO is planning an announcement on Microsoft Azure once it formalises the business partnership.
Memverge Big Memory Computing
Memverge, a supplier of in-memory computing software, has closed a $19m funding round led by Intel Capital, taking total funding to $43.5m.
The California startup today announced the Big Memory Computing scheme, a combination of DRAM, Intel Optane persistent memory and its own Memory Machine software technologies. It says it has a roadmap to achieve petabyte scale at nanosecond-speed.
Eric Burgener, research VP at IDC, said: “Without requiring any existing application rewrites, Big Memory Computing delivers the highest category of data persistence performance with the enterprise-class data services needed for the real-time, mission-critical workloads that will increasingly drive competitive differentiation for digitally transformed enterprises.”
Memverge’s technology includes instant and non-disruptive ZeroIO Snapshot for data persistence. Charles Fan, CEO and co-founder of MemVerge, said: “We founded MemVerge on the premise that every application should run in memory. There are clear use cases in AI/ML and financial market data analytics today.”
Brad Anderson, EVP of NetApp, which is a MemVerge investor, said: “MemVerge’s software combines persistent memory, hyperconverged architecture, and container orchestration, all critical components to support data-intensive applications including AI, machine learning and intelligent analytics. Our investment in MemVerge supports what we feel is the next step in the evolution of storage technology.”
MemVerge Memory Machine software will be available to select customers as part of the Early Access Program this quarter.
Nutanix improves disaster recovery
Nutanix has added new disaster recovery facilities for its AHV hypervisor:
Multi-Site Disaster Recovery: This helps customers to recover from the simultaneous failure of two or more data centres, while keeping applications and data available to users. It uses automation technology to eliminate the complexity of DR installation and on-going orchestration.
Near Sync Disaster Recovery: Nutanix now supports near sync replication with an RPO of only about 20 seconds, a 3x improvement from before.
Synchronous Data Replication is now natively supported and can be used by customers to deliver a highly available service for workloads, such as virtual desktop infrastructure, databases, and general server virtualization.
DR Orchestration with Runbooks.
Nutanix reckons it is the only leading HCI vendor to offer a 20-second RPO – Cisco, Dell EMC, VMware, HPE and NetApp can’t match it, in other words.
The new DR capabilities are included in Nutanix HCI software and are generally available.
Sony PlayStation 5 SSD
Tim Sweeney CEO of Epic, the games developer, discusses the new SSD inside Sony’s PlayStation 5 gaming console, in interview with The Verge He said it has a “multi-order bandwidth increase in storage management”.
This drive is an 825GB NVMe M.2 format SSD that holds compressed data and has 5.5GB/sec of raw data bandwidth; 9GB/sec compressed, according to TweakTown. That media outlet says this Sony SSD has a 12-channel controller and a PCIe gen 4 interface.
It will use decompressed data for real-time processing and loading. Games will boot in a second with no load screens. The drive can load 2GB of date in 0.27secs.
Sweeney said the SSD will enable a virtual game world, tens of gigabytes in size, to render on screen in less than a second because the drive can transfer information so fast: “The storage architecture on the PS5 is far ahead of anything you can buy on anything on PC for any amount of money right now. It’s going to help drive future PCs. [The PC market is] going to see this thing ship and say, ‘Oh wow, SSDs are going to need to catch up with this.”
Shorts
Copy data manager Actifio has announced its strongest Q1 to date, with record revenue and double digit Y/Y growth. It recorded several seven-figure deals from some of the hardest-hit Covid-19 regions around the globe (Northern Italy, UK, New York).
IaaS provider Priseda will offer cybersecurity-enabled private cloud DRaaS, using Asigra software, to business customers. The new GridObserver (GO) solution will be sold through the IT channel and provide full-service DRaaS and network management services.
Cloud backup and storage supplier Backblaze has provided its latest hard disk drive reliability stats in one of its regular blogs.
Hitachi Vantara, The American Heart Association, and Burst IQ& have announced a COVID-19 data challenge to accelerate understanding of the correlation between other health conditions or health disparities and COVID-19 mortality. The challenge gives researchers access to The American Heart Association’s Precision Medicine Platform, developed by Hitachi Vantara, and Burst IQ’s global COVID-19 datasets.
The US Commerce department has decided that semiconductor chips manufactured overseas using US software or hardware cannot be shipped to Huawei or any of its subsidiaries without US government permission. Wells Fargo senior analyst Aaron Rakers told subscribers: “We … believe that there is long-term potential for a technological decoupling of the U.S. and Chinese semiconductor industries, creating a worldwide bifurcation of semi technology.”
B&F suggests this means non-X86 standard CPUs could emerge from China along with separate DRAM and NAND technology. US-operated semi fabs in China, such as Intel’s Dalian facility, could be the target of any Chinese response to the US Commerce department move.
The OpenFabrics Alliance (OFA) and the Gen-Z Consortium today announced a Memorandum of Understanding (MOU) to advance industry standardisation of open-source fabric management. The MOU creates the framework for a series of technical exchange meetings between the two organisations to determine future activities. In April the CXL and Gen-Z groups announced a memorandum of understanding, which opened the door for their future collaboration.
MariaDB today announced the immediate availability of its database-as-a-service MariaDB SkySQL through the Google Cloud Marketplace. SkySQL has a cloud-native, Kubernetes-based architecture.
Nvidia’s Mellanox unit has launched the ConnectX-6 Lx SmartNIC, a 25/50 gigabit per second (Gb/s) Ethernet smart network interface controller (SmartNIC.) It uses software-defined, hardware-accelerated engines to offload security and network processing from host CPUs. The card provides RDMA over converged Ethernet (RoCE), advanced virtualization and containerisation, and NVMe over Fabrics functionality.
The chip incorporates IPsec in-line cryptography, hardware Root of Trust, and a 10x performance improvement for Connection Tracking. ConnectX-6 Lx is sampling now, with general availability expected in Q3 2020.
Portshift, which supplies cloud-native workload protection, announced it has received Red Hat’s Container certification. This expands the availability of Portshift’s workload protection across Red Hat’s hybrid cloud portfolio, including Red Hat OpenShift, to help provide more secure and compliant containerised application deployments.
OSNEXUS, a supplier of grid-scale software-defined storage, has released QuantaStor 5.6. New features include NAS storage-tiering to cloud platforms, S3 Reverse Proxy capabilities for IBM Cloud Direct Link, scale-out Ceph technology upgrades, and enhanced hardware enclosure visualisation capabilities for Ceph.
Quest Software has announced NetVault v13 to help organisations simplify data protection and recovery strategies across cloud environments and workloads. The software includes a new Rest API interface and support for Office 365 Teams.
Redis Labs has teamed up Microsoft to deliver Redis Enterprise as new, fully integrated tiers of Azure Cache for Redis. The service is available today in Private Preview. The functionality includes Redis on Flash, modules, and, in the future, the ability to create an active geo-redundant cache for hybrid-cloud architectures.
Seagate has announced a limited time offer for buyers of its PhotoDrive backup device at Walmart stores. It will include a complimentary, three-year Mylio Create plan. This includes automatic multi-device syncing for up to 50,000 photos, editing features, and cloud-optional backup storage.
SIOS has announced general availability of its v9.5 Protection Suite for Linux. This software is aimed at complex high-availability SAS S/4HANA environments and provides application-aware monitoring, cloning, and failover orchestration. It can be used on-premises and by MSPs and public cloud providers. V9.5 includes support for the latest Linux operating systems and databases, including RHEL 8.1, CentOS 8.0, CentOS 8.1, Oracle 8.0, and Oracle 8.1.
Tape library and unstructured data storage suppler Spectra Logic has launched a Remote Installation Programme to support customers during the global coronavirus outbreak. This is for customers who are restricted from having outside personnel step onsite, enabling them to receive remote assistance with installation, set-up, testing and service.
Syncsort has rebranded itself as Precisely, following its December 2019 acquisition of the Pitney Bowes software and data business. Precisely says it is a world-class business with leading software and data enrichment products, annual revenue in excess of $600m, 2,000 employees, and 12,000 customers in more than 100 countries, including 90 of the Fortune 100.
CEO Josh Rogers said: “Put simply, better data means better decisions, and Precisely offers the industry’s most complete portfolio of data integrity products, providing the trusted link between data sources and analytics that helps companies realize the value of their data and investments.”
Google has announced its Google Cloud VMware Engine, whereby customers can run applications on VMware Cloud Foundation in Google Cloud. The GC VMware engine will be generally available this quarter. DR supplier Zerto supports the GC VMware Engine and says customers can use Google Cloud as a replacement for their Disaster Recovery site. It says it provides 24×7 availability with its continuous data protection and fast RTOs (minutes) and RPO (seconds).
People
Database copy manager and data ops supplier Delphix has hired Alex Hesterberg as chief customer officer, and Deron Miller as SVP of Americas Field Operations. Hesterberg, ex-Riverbed and Pure Storage, will be responsible for Delphix’s solution consulting, implementation, pre-sales, customer success, support, education, and other services functions.
NetApp CEO George Kurian has hired ex-Microsoft exec Cesar Cernuda to be NetApp’s President, a role formerly held by Kurian. Cernuda will be responsible for NetApp’s global world-wide go-to-market operations including sales, marketing, services, and support. At Microsoft Cernuda was president of Microsoft Latin America (and a Corporate VP and looked after Microsoft’s products, services, and support offerings across Latin America.
There has been speculation that Kurian is about to leave and Cernuda is his replacement. ESG founder Steve Duplessie thinks this is wrong. Before Kurian became NetApp CEO the job was carried out by Tom Georgens. He had previously been NetApp’s President for several years, reporting to the then-CEO Dan Warmenhoven.
Nutanix has promoted Christian Alvarez to SVP w-w Channels.
Quest Software has appointed Patrick Nichols, ex-CEO at Corel, as its new CEO. His predecessor Mike Kohlsdorf is joining Francisco Partners Consulting as President.
Scale-out file storage supplier Qumulo has appointed Craig Bumpus as its chief revenue officer. Previous incumbent Eric Stollard has moved into an executive advisory role for Qumulo for the next year. Berry Russell has been appointed SVP and GM of cloud and Michael Cornwell as CTO.
Datera is cutting staff and reducing some salaries to reduce cash burn. The storage software startup last week raised an undisclosed sum in a funding round, CEO Guy Churchward wrote on LinkedIn, which involves “a plan-pivot from ‘foot jammed on gas’ to ‘get to cash flow positive by the end of FY2020”.
Datera said the reorganisation is its response to the covid-19 pandemic and its effect on the storage market.
Blocks & Files understands the company has laid off 10-15 per cent of the workforce and that Churchward has taken a temporary 80 per cent salary cut.
Last week Datera’s website stopped working briefly and CTO Hal Woods posted an ‘I’ve been laid off‘ message on LinkedIn. This prompted speculation on Reddit that Datera was in trouble. Churchward posted this response on the website.
1. I’ve been the operating CEO of Datera since late 2018 and I am still here.
2. Website had a brain fart yesterday due to an expired credit card on file and was back up shortly after.
3. To the funding point we just concluded another round with a plan-pivot from ‘foot jammed on gas‘ to ‘get to cash flow positive by the end of FY2020’. (of which I posted last week – LinkedIn update Post on Q1 and beyond .)
Even coming off a stellar 2019 result followed recently by >150% growth in Q1 bookings, we felt we should (like many others) be cautious in our forward planning.
As such we ratcheted back expenses that were being incurred to prime a faster/riskier growth trajectory and took a very small reduction in workforce mainly aimed at pairing back executive coverage to the fundamentals of customers and product support.
Every company is adapting and flexing to the current market conditions, I could already see CFBE (Cash Flow Break Even) on the near horizon, so a plan adjustment to accomplish this in 2020 would be an epic achievement. This without question would make Datera stronger for future growth and open many more avenues and opportunities to us.
Update: Two other execs are no longer listed on Datera’s leadership webpage;
Claudio Fleiner, Co-founder and Chief Architect,
Chris Cummings, CMO.
Datera supplies scale-out enterprise-class clustered block and object storage software as an alternative to hyperconverged infrastructure (HCI) and traditional external SAN array storage. It has a reselling partnership with HPE, with its software part of HPE’s Complete program.
The company reported 325 per cent revenue growth in 2019 compared to 2018, and has transitioned from perpetual license sales to a subscription business model.
Update: May 18 – TouchSupport customer review of ExaFlash One.
Nimbus Data, a data storage vendor based in Los Angeles, has announced the ExaFlash One all-flash array with a multi-function, multi-protocol, multi-mode OS, and simplified and transparent pricing.
The company has recruited Eric Burgener, research VP in IDC’s Infrastructure Systems, Platforms and Technologies Group, for this quote: “With its versatility, federated architecture, and multi-tenant management capabilities, Nimbus Data AFX is a well-conceived solid state storage platform for dense enterprise workload consolidation.”
A customer, Nick Smith, director of Remote IT and Broadcast Engineering at World Wrestling Entertainment, said: “We were able to deploy it in 30 minutes. We are consistently writing to the ExaFlash array at line-rate, and it is performing flawlessly.”
The array hardware is a 2U x 24 x U.2 slot enclosure with dual active:active controllers.
ExaFlash One
Supported host links are 8 x 16Gbit/s FC, 8x 10 or 40GbitE, 8 x InfiniBand FDR, or 4 x 16Gbit/s FC + 4 x 10GbitE. A maximum 2 million IOPS of performance is claimed.
The case is populated with 48TB (12 x 4TB SSDs) to 768 TB (24 x 32TB SSDs) or delivered empty with a set of qualified 12Gbit/s SAS SSDs from Kioxia, Samsung, Seagate and Western Digital that customers can buy themselves.
The arrays can be federated but there is no back-end fabric, and no data IO between or across the federated systems. This is not clustering – the federation is via scale-out independent nodes.
ExaFlash One has a separate out-of-band Ethernet management link with a single and central point of management.
AFX software
The AFX operating system is claimed to provide block, file and object storage (S3) access with a wide set of supported access protocols:
It is also said to include data management features including snapshots, cloning, replication, multi-pathing, RAID, thin provisioning, compression, deduplication, checksums and encryption.
The array can be operated in three modes;
Intelligent mode – with the data management features
Ludicrous mode – RAID protection only and optimised for throughput and IOPS
Bare-metal mode – with direct access to each SSD using NVMe-oF, iSCSI. FC or SRP
Support and pricing
Nimbus has announced a support scheme with a flat annual fee per array and a single support tier with 24 x 7 x 365 availability and next-day parts replacement. There are free non-disruptive controller upgrades after three years, all-inclusive software, free software updates and an absence of capacity-based license fees.
The company said the array’s power usage is carbon-neutral, courtesy of a renewable energy offset scheme. There is a 60-day money-back guarantee and support contracts run from one to seven years.
Customers can set up an account on the Nimbus website, spec out and price an array themselves. Nimbus claims a $10,000 starting price but we were only able to get down to $20,000 using the online facility.
Update: That was by selecting the ‘No SSDs included’ option. Nimbus explains that $10,000 is the ExaFlash One system base price, including all software, and Tectonic is $10,000 per year.
Choosing the maximum capacity system of 768TB would cost $212,000.
TouchSupport customer view
Touch Support CTO Jason Hamilton was quoted in Nimbus’ announcement; “We chose ExaFlash One for our internal VMware environment and have been very pleased with its speed, availability, and ease of use. Given our experience, we recommend ExaFlash to our clients for their critical workloads.”
Touch Support is based in Chicago. We contacted Jason and asked him three questions about his decision to buy the ExaFlash One array.
Blocks & Files: Could you tell me what were the alternative arrays you considered for your workload and why you chose the ExaFlash ONE?
Jason Hamilton: When we started research for a new SAN I cast a wide net but ultimately narrowed our options down to three choices. HPE Nimble all-flash, Fujitsu (ETERNUS AF250 S2) and of course Nimbus ExaFlash.
There were a couple of reasons that we chose Nimbus. Firstly and probably most was price. Nimbus is priced very competitively even factoring in a 5-year support contract paid upfront.
Next, it’s important to know that I’ve been following Nimbus since mid-2012 where I learned of the company and their S-series arrays at a VMware conference. Even back then they were producing all-flash arrays touting 1 million IOPS. This impressed me greatly and I kept an eye on their progress over the years.
The Fujitsu option is very much a new player in the segment and didn’t give me as much confidence in their ability to maintain the product considering their diversified nature.
Lastly, going with ExaFlash seemed to make the most sense. This is their most recent flagship platform and would stand to have the longest support-life period available. We generally try to maximize our hardware lifecycle and it is critical to maintain an active support contract on critical systems such as a SAN.
Blocks & Files: How are you using the array? I mean for block, file or object work?
Jason Hamilton: We are an all VMware shop so the array is configured only for iSCSI block access to our ESXi hosts. We naturally do use the compression and deduplication features at the array level as well.
Blocks & Files: How does it compare to other arrays you have had experience with?
Jason Hamilton: Having either closely worked with or at the very least run proof-of-concept on a number of SAN and NAS solutions, the Nimbus ExaFlash is refreshingly simple. Coming from a Unix/Linux background the option to handle all configurations completely from an ssh shell is very pleasing. The web-based interface is likewise very simple to use and doesn’t have any irksome custom plugin requirements. Gone are the days of separate fat-client software or the hit-or-miss vSphere plugins for management and configuration.
Background
Nimbus has sold ExaFlash arrays since 2016 and claimed 200 customers in 2018. It now says it has 600 deployments. CEO and founder Thomas Isakovich says Nimbus was profitable in 2019 and expects to grow profits this year.
The privately-owned company appears to be very small. Six employees are listed for Nimbus Data on LinkedIn and the leadership page on its website lists one executive officer – Isakovich.
We already knew Liqid’s PCIe 4.0 SSD was fast – and now Liqid with Dell help has proved it.
Liqid makes composable systems using a PCIe fabric to link CPUs, GPUs, storage and networking server elements. It stuffed the company’s LQD4500 PCIe gen 4 add-in-card SSD inside a Dll PowerEdge R7515 rackmount server and produced a wicked fast box.
The R7515 is a single socket server in a 2U cab. It uses an AMD EPYC 7742 CPU with 64 cores and 128 PCIe gen 4 lanes, with a base 2.25GHz clock rate boostable to 3.4GHz, and up to 4TB of DRAM.
Dell PowerEdge R7515
Dell’s R7515 also has a couple of PCIe gen 4 slots, one of which can be occupied by a Liqid LQD4500 add-in-card SSD. It has an NVMe interface and 30TB of capacity.
Liqid ran a R7515/LQD4500 system through its paces at its Colorado headquarters, with performance tests for random and sequential IO. The test system delivered 3.14 million random read IOPs (4K block size) and 961,663 random write IOPS. With 8K blocks, the benchmarks were 1.8 million random read and 479,889 random write IOPS.
Dell R7515 + Liqid LQ4500 IOPS results
Sequential bandwidth with 128K blocks was 24.471GB/sec reads and 10.052GB/sec writes. It was about the same with 256K blocks – 24.481GB/sec reads and 10.028GB/sec writes.
Socket to me
Liqid claims this R751 “powered by AMD and Liqid enables users to deploy server systems with the highest compute and storage performance on the market. This leading edge Gen4 PCIe platform is enabling next levels of performance not previously available from a single server let alone a single socket server.”
You can check out the case study for more details. PCIe gen 4 systems are going to reset everyone’s PCIe gen 3-based performance clock.
Enterprise SSD sales are rising sharply but so too are nearline disk drive sales, according to a TrendFocus SSD industry review of the first 2020 quarter. Nearline disks are ten per cent of the capacity cost of SSDs – which is the very simple reason for the robust performance of this storage media format.
About 12.7 million enterprise SSDs shipped in the quarter, totalling 20EB in capacity – a 134 per cent capacity jump year over year (Y/Y). Even so, enterprise SSD capacity shipments amount to around 10 per cent of the HDD capacity shipped in the quarter, and HDD capacity shipped was up 94 per cent Y/Y.
The chart above, by Wells Fargo analyst Aaron Rakers, shows enterprise HDD capacity shipped rising more steeply than enterprise SSDs from 2012 onwards. The curve grew more steeply in 2019. Enterprise SSD capacity shipped also grew more quickly in 2019 but at a slower rate than HDDs.
Rakers tells subscribers: “On a $/TB basis, we estimate that enterprise SSDs are approximately 10x more expensive than nearline enterprise HDDs.”
Another Rakers-produced chart, shown below, plots the $/TB price differential between enterprise SSDs and nearline disk drives. It plummets in 2013, falls at a slower rate in 2014, slower still in 2015-2018 and levelled off in 2019. The blue bars in the chart show the SSDs at nine to 10 x the nearline HDD price in 2019 and 2020’s first quarter.
More layers for 3D NAND and energy-assisted recording for HDDs will soon hit the market. These capacity advancements should help maintain the cost differentials between the two media types.
SSDs are about 2.2x more expensive on a $/TB basis than mission-critical HDDs – the 10,000rpm 2.5-inch disk drives. This helps explain why all-flash arrays are supplanting disk-based arrays for storing primary enterprise data.
The TrendFocus study shows that the category of enterprise SSDs using the PCIe/NVMe interface accounted for just over 11EB of capacity, nearly tripling Y/Y, and 55 per cent of total capacity shipped.
Rakers comments: “We continue to believe enterprise PCIe/NVMe SSDs will grow to account for the majority of the enterprise SSD capacity shipped over the next few years.”
The approximate supplier enterprise SSD capacity shipped shares in the quarter were:
Samsung – 42%
Intel – 23%
Micron – 10%
Western Digital – 7%
Those shares are different if we just look at PCIe/NVMe enterprise SSDs;
Samsung – 42%
Intel – 28%
SK hynix – 11%
Western Digital – 9%
Micron – 3%
Kioxia – 2%
There’s more competition and Micron needs to up its game.
PCIe 4.0 SSDs, that are double PCIe gen 3’s speed, are coming. These will help spread the NVMe message, and widen the access speed difference with mission-critical disk drives. Blocks & Files thinks PCIe gen 4 could be the kiss of death for this class of disk drive.
The faster PCIe interface could also make QLC (quad-level cell) SSDs more attractive for fast access archive data. The future is bright for NAND foundries and SSD manufacturers. But in the absence of technology development disasters, the future for nearline disk drives is also bright – because SSD capacity technology can’t overtake disk drive technology.
Veeam, the backup software vendor, has announced a blow-out first 2020 quarter, with customer count passing 375,000 and annual recurring revenue (ARR) growing 21 per cent.
Veeam CEO Bill Largent said in a statement today: “The first few months of the year have been unusual to say the least for all of us. We are having to conduct business in a different way, ensuring that the Veeam community – employees, partners and customers – stay safe during this period. I’m very proud of our 4,300 employees as we have continued to grow during this challenging time while not missing a beat on customer support.”
Veeam reports 97 per cent year-on-year (Y/Y) growth in Veeam Universal License subscriptions. And it pointed to the IDC Semi-Annual Software Tracker, 2019H2, which records the fastest Y/Y revenue growth for Veeam – at 20.5 per cent – among the top five data protection vendors.
This matches the company’s performance in the previous edition of the IDC tracker, when it had 22.4 per cent Y/Y growth for the first half of 2019 and estimated revenues of $427m. The company said in May 2019 that it had hit a billion dollar revenue run rate.
Veeam was bought by the private equity firm Insight Partners for about $5bn earlier this year.
StorONE CEO Gal Naor today challenged storage array vendors to reveal their prices in public. “Buying an enterprise storage system shouldn’t be as hard as buying a car, but should be as easy as buying a smartphone,” he said.
The software-defined storage startup has built an online configurator with openly available pricing for its TRU (Total Resource Utilisation) S1 software, engineering it for efficiency and speed. The software supports block (Fibre, iSCSI), file (NFS, SMB) and object (S3) use cases and runs on industry standard servers from Dell, HPE and Supermicro.
“S1:TRU price simplifies the storage procurement process,” Naor said. “No more back and forth haggling over solution costs. No more wondering if you would have gotten a better deal at the end of the quarter. S1:TRUprice allows enterprises to get the best pricing, upfront, while also getting the best performance and data resiliency features they need.”
StorONE licenses S1 software in either 1, 3 or 5-year subscriptions. When a user selects a system and confirms purchase, StorONE ships the whole system to the customer for contactless, remote installation and training. And it thinks it has a big price advantage over its rivals.
George Crump, StorONE’s chief marketing officer, said customers “can come to our site, price out their exact solution and compare it to what they are paying now. In most cases, we are confident they will find that the 3-year TCO on our solution will be less than what they are paying for one year of maintenance with their current vendor.”
Playing around with StorONE pricing configurator.
Interested people can trek over to the StorONE website’s pricing section where they can use a three-year TCO pricing configurator without registering their details, and with instant prices as they configure a system’s server, availability, media, capacity and networking cards.
We tried it out. A 736TB high-availability all flash HPE-based system had a $346,318.14 3-year TCO. Changing to a Supermicro server moved the price to $341,860.78. A 1PB standard (not HA) all-disk Supermicro server system would cost $138,36.00. It’s fun playing around with the options and easier to use than the Nimbus Data ExaFlash One pricing configurator, which we also tried out this week.
Will Dell EMC, HPE, Hitachi Vantara, IBM, NetApp and Pure Storage publish their prices? We all know the answer to that.
