Home Blog Page 434

Teradata moves to new Vantage point as it advises customers to ‘stop buying analytics’

By
Chris Mellor
-

Teradata is exhibiting many signs of a mid-life crisis. The data warehouse vendor has upped sticks, moving its corporate headquarters from Dayton, Ohio to San Diego, given its logo a corporate makeover and adopted a new name for its flagship product.

It also has a new mission statement: “Stop buying analytics and start investing in answers.”  This is ballsy for a company that makes its living from selling analytics but makes more sense when you read the rest of the statement: “With all of the investment being made, there’s one question no one seems to be asking: When do we stop buying into partial solutions that overpromise and underdeliver? The answer to that question is ‘now.’”

What gives? and What is “now”?

Black and lower case is the new orange and upper case

Compare and Contrast with the old logo. 

Burnt orange

‘Now” is the company’s new flagship product, Vantage – the new name for the Teradata Analytics Platform, announced in October 2017. The idea is to simplify the purchase of many disparate products, and it integrates Teradata and Aster technology with  third party tools and analytical engines.  These include Spark, TensorFlow, Gluon and Theano, a range of algorithms for deep learning, and scalable analytic functions.

Vantage is deployable in the Teradata Cloud or on its hardware, in the public cloud or run on VMware-powered commodity hardware, with portable and subscription-based licences.

Vantage is called a pervasive analytics platform, and continues the integration story with more than 180 pre-built analytic functions and engines. It provides access to descriptive, predictive and prescriptive analytics and includes autonomous decision making, machine learning functions and visualisation tools.

There is a native graph processing engine to help identify and measure relationships between people, products and processes.

Teradata says users can “switch between the most common interfaces and tools, including SQL, Python and R, as well as a broad set of BI and visualization tools, with the advanced functionality of SAS, Jupyter and RStudio. The platform also provides storage and analysis for multi-structured data, including JSON, BSON, AVRO, CSV, and XML.”

Teradata’s Vantage is available now, along with more than 5,000 Teradata experts for service help, and there is a downloadable datasheet. 

Egnyte thinks global with $75m injection from Goldman Sachs

By
Chris Mellor
-

Egnyte has secured $75m in a new funding round – five years after its last round and 18 months after its CEO said he didn’t need more cash.

The rationale for taking the funding now is to get Egnyte operating on a global scale. In April 2017 Vineet Jain, Egnyte CEO, told me that he would consider raising more money “if we can add another 14-15 points to the growth curve.”

It looks like he’s convinced of that now. In a statement Jain said the investment will “allow us to focus on the long-term goals for our business – increasing our global presence, expanding our product footprint, and delivering the best customer experience in the industry.”

Egnyte co-founder and CEO Vineet Jain

Egnyte is a cloud storage and file sharing vendor whose competitors include Dropbox and Box Inc. The company helps businesses migrate their infrastructure to the public cloud by replacing file servers, improving content management systems, and protecting sensitive content.

The firm says it has been cashflow positive since 2016 and has recorded more than eight quarters of positive growth. It claims more than 14,000 paying customers worldwide including Arista, Balfour Beatty, BuzzFeed, Nasdaq, Red Bull, and Yamaha.

The E-round was funded by Goldman Sachs and brings Egnyte’s total funding to $132.5m.

You’re an AI startup? Micron has $100m up for grabs

By
Chris Mellor
-

Artificial intelligence is a huge money opportunity for memory and storage vendors, according to Micron boss Sanjay Mehrotra. And he is putting the company’s money where his mouth is.

The semiconductor giant will through its investment arm Micron Ventures spend up to $100m in AI technology startups, of which $20m is set aside for startups led by women and other under-represented groups.

The company made the announcement yesterday at the inaugural Micron Insight event in San Francisco, which focused on how memory technology enables AI, machine learning (ML) and data science technology.

AI and machine learning systems tend to involve servers with larger than average memory amounts and lots of flash memory storage. Micron President and CEO Sanjay Mehrotra has a canned quote about this: “These (AI, ML and deep learning) trends are at the heart of the biggest opportunities in front of us, and increasingly require memory and storage technologies to turn vast amounts of data into insights that accelerate intelligence.”

Research grants

In related news The Micron Foundation has launched a $1 million grant for universities and non-profit organisations to conduct research into AI.

The  fund is available to select research universities focused on the future implications of AI in life, healthcare and business

Three initial recipients get half a million bucks between them.

  • The non-profit A14AII works to increase diversity and inclusion in AI education, research, development and policy, with activities such its AI Summer Camp which is open to 9th– 11th grade students
  • Berkeley Artificial Intelligence Research (BAIR) Lab which supports researchers and graduate students developing fundamental advances in computer vision, machine learning, natural-language processing, planning and robotics 
  • Stanford Precision Health and Integrated Diagnostics Center which funds researchers focused on testing and disseminating the next generation of science and technology to deliver predictive, preventive and precise care to every patient


SAP slaps sticker on Lenovo HANA HCI box

By
Chris Mellor
-

SAP has given its seal of approval to a Lenovo hyperconverged appliance using Nutanix AHV virtualization software to run its HANA in-memory database.

The ThinkAgile HX Solution for SAP HANA, to use its full name, includes onsite deployment of SAP HANA and full SAP solution support by Lenovo.

The ThinkAgile system is preloaded with Nutanix software in the Lenovo factory where it is integrated and validated with firmware.

Lenovo HX7820

It runs the Nutanix Enterprise Cloud OS, meaning the AHV hypervisor  and is configurable with SAP HANA-compatible options only .

The hardware spec includes a 4-socket server that has 4 x Intel Xeon Gold or Platinum processors, and up to 24 x 2.5-inch drive bays in its 4U enclosure.


Memory prices are on their way down

By
Chris Mellor
-

Trendforce forecasts a five per cent or greater fall in DRAM contract prices in 2018’s fourth quarter, following 1 to 2 per cent rise in the third quarter. The expected price fall will bring to an end to nine consecutive quarters of price rises. NAND prices are expected to fall even more steeply.

Trendforce notes soft demand for server DRAM and  thinks average DRAM prices could drop 15-20 per cent in 2019. The price drop could be even steeper if demand for servers and smartphones weakens further, the market research firm says.

It attributes the DRAM to weak smartphone shipments, low server demand and PC and notebook demand squeezed by Intel CPU shortages. These three factors could lead to DRAM over-supply in 2019.

Trendforce expects 2019 annual bit output to increase by nearly 22 per cent, with the 1X/1Y nm processes becoming mature and wafer starts increasing, although manufacturers have become more conservative in capacity expansion.

NAND

NAND prices have already started falling, with a 10 per cent decline in the third quarter and a further 10 -15 per cent fall is expected in the fourth quarter. 3D TLC (3 bits/cell) chip contract prices may fall even more steeply in the fourth quarter.

Micron NAND wafer.

For 2019 a potential 25-30 per cent price fall is on the cards due to increased 3D NAND production capacity. Trendforce notes sluggish demand for flash in the consumer electronics sector and strong demand for enterprise SSDs. In the latter case, vendor competition and improved yield of 64-layer and 72-layer 3D NAND chips from wafer starts will put a crimp in prices.

Demand in the first half of 2019 will be seasonally weak and could be affected by the US-China trade war. Trendforce said suppliers could moderate their transition to higher output 96-layer and estimates overall NAND production capacity of NAND will grow 5 per cent annually by the fourth 2019 quarter. It expects memory makers to cut output spending budgets.

Show me the money! SPEC SFS 2014 benchmark is not fit for purpose

By
Chris Mellor
-

Opinion: The SPEC SFS 2014 benchmark shows which system vendor are fastest at a set of file system performance tasks, but there are no price/performance numbers. This means you either win the benchmark or you don’t and there is no opportunity for showing the relative value of your result.

The SPEC SFS 2014 benchmark resolves the shortcomings of an earlier version,  which meant that you could not compare NFS and SMB/CIFS results. It is also an end-to-end system and not a component-level test. For example, it looks at disk-level performance rather than end-to-end system level performance.

There are five subtests:

  • Number of simultaneous builds that can be done (software Builds)
  • Number of video streams that can be captured (VDA)
  • Number of simultaneous databases that can be sustained (Database)
  • Number of virtual desktops that can be maintained (VDI)
  • Electric Design Automation workload (EDA)

So far, vendors have runs several software build tests but have shied away from others such as VDA, Database, VDI and EDA. Effectively,  this is an all-or-nothing software build test where you aim score the highest number of builds and the lowest overall response time (ORT).

There are sub tests at varying software build levels but the basic idea is still to complete as many builds as possible in the shortest possible time.

Sample SPEC SFS 2014 SW build results

For example, in the chart above the Oracle ZFS system could be one quarter the cost of higher-scoring systems. If so that would make it a serious price/performance contender. But the benchmark gives us no means of telling.

SPC-1 IOPS benchmark with price/performance attributes

Compare this with the SPC-1 benchmark which looks at a storage array’s response to IO requests and  has a price/performance element.

That means that you can win the top SPC-1 IOPS results but also you can win the best results in a system price range – $1m to $2m, for example. That gives smaller and mid-range suppliers valid reasons to use the SPC-1 IOPS benchmark.  They have no fear of getting trampled by a massive competitor with, say, a $4m+ array throwing SSDs and controllers at the storage IOPS challenge to batter it into submission.

To summarise, SPEC SFS 2014 is needlessly restricted to an all-nothing high score/low ORT niche. Customers need a benchmark that can help them assess the gamut of suppliers, large and small.

Superfast filer WekaIO recruits filer tech guru

By
Chris Mellor
-

Andy Watson has left Minio, an object storage software company where he was one of three CTOs, to take up the same role at WekaIO. His new company is an Israeli startup that has developed its own file storage stack to speed file access. It scores high on the SPEC SFS 2014 benchmark.

Andy Watson, now one of one CTOs instead of being one of three.

Watson is a long-time filer guy, and his CV includes stints at EMC, NetApp, and  – for filer archaeologists – Auspex Systems. In his new job Watson will guide technology development for vertical markets, monitor emerging trends and communicate them to key business stakeholders.

Your Irregular Storage Digest with Delphix, HTF Market Intelligence, Mimecast, and Korean fabbers

By
Chris Mellor
-

Here’s the speed read version: Delphix has picked up a UK TV broadcaster as a customer; a somewhat limited cloud backup market report appears; Mimecast is losing its CFO; and Samsung and SK Hynix are expanding their semiconductor fab lines.

Delphix

Channel 4,  the UK TV broadcaster, has become a Delphix customer, and says its application release cycles will shorten as a result.

Delphix, a database virtualizer, will be involved with Channel 4’s broadcast management and advertising scheduling platforms. Channel 4 makes regular updates to these systems, and its test and dev people need test data to check the changes out.

It says the time taken to refresh the 18TB or so of data in test environments was slowing down  system and business development.

The company started using Delphix Dynamic Data Platform software to automate data delivery to test and dev, taking advantage of Delphix Data Pods to connect database administrators to developers to ensure faster delivery of data. What used to take days now takes just minutes, Delphix says, enabling Channel 4 to release applications more quickly.

The broadcaster also has a suite of data warehouses to provide insights into business performance and indicate future opportunities. Its using Delphix software to refresh this information and manage volume sizes better.

Delphix digression

Delphix founder and exec chairman Jedidiah Yueh, has written a book, Disrupt or Die, about the need for disruptive startups and the effects and characteristics of Silicon Valley funding, assumptions and development on them.

Before starting up Delphix he founded, developed and sold deduplicating backup supplier Avamar to EMC, for $165m in November 2006.

According to Yeah, Avamar, based in Orange County, was not in an optimum location, in comparison to Silicon Valley-based Data Domain, its main competitor. That made it harder get top engineering and executive talent and: “our engineering team shipped product too late. We had a commanding two-year lead, but a Silicon Valley competitor ran us down.” 

That was Data Domain.

“In the end we sold for $165m, and our competitor sold for $2.4bn.”

Lesson learnt; Delphix moved to Silicon Valley in 2009.

 Yueh’s book is full of interesting insights and viewpoints and is available on Amazon.

Global cloud backup market report

A report by HTF Market Intelligence said: “The global cloud backup market size was $1.2bn in 2017 and is expected to reach $6.82bn by the end of 2025, with a CAGR of 23.4 per cent during 2018-2025.”

HFT Market Intelligence is not highly-specialised from an IT market point of view and has produced report on t global table and kitchen glassware, the global zipper market and a global yoga mat study.

The cloud backup report looks at Acronis, Asigra, Barracuda Networks, Carbonite, Code42, Datto, Druva, Efolder, IBM, Iron Mountain, Microsoft and Veeam Software.

Missing vendors include Arcserve, Commvault, Dell EMC, HPE, Kaseya, Rubrik, and Veritas.

The cloud storage report costs $3,300 for a single user and $6,600 for an enterprise copy.

If you request a PDF sample copy  you get a reply message saying: ”Thank you for your enquiry. Our Executive will contact you soon.”

And they will. I was phoned within five minutes on a Saturday morning. The sample copy offer seems to be a sales lead generation exercise, containing next to no useful supplier information at all. Bargepoles and long spoons come to mind.

Mimecast losing CFO

Peter Campbell, the CFO of email security and backup and archiving supplier Mimecast, is  resigning, effective March 31, 2019. There will be a se-month transition period in which he stays on while Mimecast looks for a successor. An SEC filing revealing this was made last month.

Mimecast co-founder and CEO Peter Bauer, issued a warm canned quote about what seems to be an amicable departure.

“Peter has been an exceptional leader at Mimecast for the last 12 years, overseeing the company in its early years, leading the company’s global expansion, and guiding the company through its initial public offering and life as a public company. He built a world-class, global finance organization, and played a critical role in developing and executing the company’s growth strategy. We thank him for his countless contributions and his assistance throughout the upcoming transition process.”

There is no word on why Campbell is leaving or where he is going.

Samsung and SK Hynix fab build-out

 Samsung and SK Hynix are building at least five new fab lines, Wells Fargo senior analyst Aaron Rakers reports.

  • Samsung is expanding its production capacity in Pyeongtaek with a second line for its next-generation 64-Layer V-NAND. That should be ready in 2020.
  • It’s adding another line to its Xian, China, fab, with mass production starting in 2019.
  • Samsung is deploying EUV tooling in its Hwaseong (Gyeonggi Provice), South Korea facility, with mass production also slated for 2019.

SK Hynix is adding new semi-conductor fab lines to three facilities:

  • Its M15 fab in Cheongju, South Korea, with a budget of 2.2 trillion Won ($1.96bn)
  • Building a new facility in Wuxi, China, ready for the end of this year
  • Building an M16 fab in Icheon, South Korea, with completion by October 2020.

We should see expanded DRAM and NAND production next year with a further increase in 2020. This lends credence to fears of an impending NAND (and DRAM) over-supply and consequent price collapse.


Exagrid exceeds quarterly revenue record

By
Chris Mellor
-

Dedupe to disk backup target array supplier Exagrid notched up 30 per cent growth in the third quarter of 2018, taking revenue to a record. But a record what?

As a rule, we ignore press releases from privately-held startups touting their revenue growth figures – as they never say what their sales actually are. But for once we will make an exception for Exagrid, even though it has done that very annoying thing – trumpeted sales growth from an invisible number.

FWIW, we think the company’s revenues currently exceed $50m and we can deduce some interesting facts from its press release. In July 2018 Exagrid said it was expanding its worldwide sales capability, and hinted at a 2019 IPO.  We infer that this is on track for sometime in 2020.

Although Exagrid doesn’t say what its revenues are,  it has revealed some previous quarterly growth rates:

  • 22 per cent revenue growth for 2018’s second quarter
  • 20 per cent in the first 2018 quarter 
  • 20 per cent in the final 2017 quarter

And we have 30 per cent in the third quarter.

We’ve asked the company if it can say more about its revenue number. In response CEO and President Bill Andrews told us via email: “We don’t disclose our revenue. We sell against Dell, HP, NTAP, IBM, Commvault, Veritas so any number would look small compared to those players.”

Cash burn? What cash burn!

But he says the company has 2,400 unique customers and more than 10,000 Exagrid systems installed worldwide. And he notes the company is gaining market share at the expense of its competitors and will end calendar year 2018 with 22 per cent growth over last year. Exagrid is gunning for 30 per cent growth in 2019,  according to Andrews.

Also, the company appears to be profitable or at least not losing too much money.  “Unlike other high tech companies, we are not burning cash and in fact we have not raised money in over 7 years,” Andrews says. “We are not a Tintri, Tegile, Violin, SimpliVity, Data Gravity, etc. that will run out of money and hit the wall.”

A prepared quote from Andrews said: “ExaGrid now offers the largest system in the industry – one that can take in up to a 2PB full backup at up to 432TB/hr., which is 3X faster than its closest competitor,” understood to be Dell EMC’s Data Domain.

Andrews tells us Exagrid’s product is: “3 times faster for ingest over Dell Data Domain, 20 times faster for restores over Dell Data Domain [and] half the price of Dell Data Domain in over 100TB accounts, which is where we focus.”

We await an upcoming Gartner magic quadrant for data centre backup and recovery which contain some useful metrics for measuring  Exagrid’s market position. However, analysts on the Gartner  team have left the firm to join vendors. So we will maybe have to wait a little longer.

Larger Mass.

Exagrid is moving corporate offices at the end of the year, Andrews says, to “larger and much nicer space in Marlborough Massachusetts. The new space is very high tech looking, it was the previous VCE space which was the joint venture between EMC and Cisco that went belly up. As you can imagine EMC and Cisco really built out the space to look like 2020 high tech space.”

Tintri co-founders join DDN

By
Chris Mellor
-

Tintri co-founders Kieran Harty and Mark Gritter have joined DDN’s Tintri business unit.  Gritter assumes the CTO position while Harty has the more nebulous role of “full-time advisor”.

Mark Gritter.

In a statement Gritter said: “With DDN’s broad market reach, significant financial investments, and synergistic data storage technologies, I’m inspired to lead Tintri by DDN’s engineering innovation into broader data management, analytics and hybrid cloud areas.”

DDN bought the company out of Chapter 11 bankruptcy for $60m in July 2018. It pledged to maintain all support contracts and aims to hire 100 people for its new business unit by year-end.

DDN customer win

In related news, the University of Michigan’s Advanced Research Computing – Technology Services unit has bought 5PB of DDN storage for its Locker Large-File Storage service. This includes massive raw video files and sensor data produced by Mcity, an urban test facility for connected and automated vehicles.

 Large image datasets generated by electron microscopy churns out more data. As does a university-wide Precision Health initiative that applies genomic, epigenetic and public policy research to better understand the risk factors associated with long-term opioid use and abuse. This is classic academic DDN storage use.

DDN announced its A3I systems today; storage arrays twinned with Nvidia GPU servers. With this entry into enterprise AI server/storage  and the Tintri enterprise storage array business DDN is stepping energetically  outside its supercomputing comfort zone.

Miscreants use Azure blob storage in phishing bait

By
Chris Mellor
-

Azure blob storage has been used in phishing attempts, with Microsoft-issued domain and SSL certification that makes the bait seem credible.

The exploit involves a PDF decoy document purported to belong to a Denver, CO law firm and hosted in Google Drive.  This contained a link to an Office 365 phishing page hosted in Azure blob storage.

Netskope, a cloud security vendor which uncovered the attack, detected the PDF decoy docs as PDF_PHISH.Gen1. It was described in a host email as  “Scanned Document… Please Review.pdf” – such things traditionally arrive as email attachments sent to victims.

Upon clicking the link, victims were given a ‘Download PDF’ button. Clicking on that button generates a message about downloading an Azure Blob storage document – https://onedriveunbound80343[.]blob.core.windows.net –  with a clickable hyperlink:

Top right box enlarged.

Clicking that leads the victim to a log-in screen asking them to sign in with their valid Office 365 email account. After some to-and-fro the victim is directed to a download page but nothing is actually downloaded.

Netskope found a similar phishing attack purporting to be from an Oregon-based dental equipment manufacturer. The company says this attack appears to be broadly targeted toward Office 365 users in the US.

How can users defend themselves?

Netskope says “The phishing bait was a Microsoft Office 365 login page, hosted on a Microsoft domain, with a valid Microsoft-issued certificate. Users can recognize that this is a phishing site based on the subdomain, which indicates that it is in Azure blob storage, and not an official Microsoft website. “

It recommends: “Enterprises should educate their users to recognize AWS, Azure, and GCP object store URLs, so they can discern phishing sites from official sites. “

In particular: “Always check the domain of the link. Know the domains typically used when you login to sensitive services. Additionally, be able to identify common object store domains, such as those used by Azure blob storage. This knowledge will help you differentiate between well-crafted phishing sites and official sites.”

Vice advice

It’s a Wild West phishing nightmare out there.  But it unrealistic to ask unreliable potential victims to manually check the domain of links. It is also unfeasible to teach users to doubt the reliability of Microsoft-issued domain and SSL certification. What is the point of these certifications if they can be effectively counterfeited?

The Netskopers say: “Get comprehensive threat and malware detection for IaaS, SaaS, PaaS, and the web with real-time, multi-layered threat detection and remediation to prevent your organization from unknowingly spreading similar threats.”

Yes indeed. Or do not click on hyperlinks in emails unless you are certain it’s safe to do so.

San Francisco Giants bags $15K savings with Cohesity gig

By
Chris Mellor
-

 Cohesity has knocked Veeam, BackupExec and Data Domain out of the San Francisco Giants’ AT&T Park in San Francisco.

Cohesity says the Giants saved more than $15,000 per year due to the elimination of the co-location site and lower maintenance and service costs. The baseball team also reduced backup time by 50 per cent with the new set-up.

 The San Francisco Giants is using Cohesity to backup and consolidate its data infrastructure, including video files and scouting reports, and use public cloud infrastructure behind the on-premises IT kit as a backup target.

Several years ago the Giants’ IT team moved to Pure Storage for virtualization storage with EMC Isilon for file storage, managing about 125TB of data at the time, including video, publications, photos, and archives. 

They used Veeam and Symantec Backup Exec to back up to a Data Domain appliance which replicated data to an off-site co-location facility.

The IT team reckoned the software and systems involved wouldn’t be able to cope with a large influx of data due to greater use of video and tracking technology on the field.

The Giants worked with system integrator Groupware Technology, Inc. and deployed Cohesity’s DataPlatform and DataProtect with C2300 hyperconverged nodes. Amazon Glacier is used for long-term storage and the Giants aims to deepen its use of the public cloud. Some 350TB of data is currently involved, and expected to grow.

The roadmap sees the Giants using Cohesity’s SQL clone and copy attach feature on its SQL database for its test/dev operations, with full cloning of a database to any SQL server or instance using native APIs. This will enable the baseball side of the house to reference database snapshots for troubleshooting and analysis.

Comment  A contract win in which Veeam knocks out BackupExec and Data Domain would not be a surprise, but a win which sees Cohesity knock out Veeam is a turn up for the books. The Giants are a marquee customer and the purchase details might reflect its value to Cohesity.  It will be interesting to see if the Cohesity’s answer to a Veeam/BackupExec/Data Domain and a co-lo offsite facility scheme appeals to a wider audience.

1...433434435...442Page 434 of 442

ABOUT US

Blocks & Files is a storage news, information and analysis site covering storage media, devices from drives through arrays to server-based storage, cloud storage, networking and protocols, data management, suppliers and standards. We publish news, views, opinions and analysis of storage industry events and developments, aiming to keep storage users, developers, consultants, marketeers, salespeople, support staff, analysts and everyone else involved in IT storage up to date with what’s going on.

FOLLOW US

© Situation Publishing, 2018-2024