Gotta speed up HPC arrays. Flash or disk, flash or disk. Let's do both – Seagate

Hybrid set-up plus Nytro caching software sends small I/Os to SSD


Seagate’s ClusterStor 300N storage array for high-performance computing is a hybrid flash/disk array, lifting sequential bandwidth by 33 per cent and small, random I/O by up to 1,200 per cent.

The idea is that the 300N systems run mixed workloads much better than previous ClusterStor arrays, by using flash to accelerate IOs that disk handles relatively badly. Seagate has added SSDs to the media mix plus software to handle them in conjunction with the disks. It claims applications will see the up to 12x IO speed benefit from SSDs for small I/O at a cost slightly above an HDD-only approach.

Thus should help ClusterStor compete more effectively against arrays such as those from DataDirect Networks, with that company’s burst buffer technology. A canned quote from Ken Claffey, Seagate VP and general manager of its HPC systems business, says: “Seagate’s ClusterStor 300N expands on our proven, engineered systems approach that delivers performance efficiency and value for HPC environments of any size, using a hybrid technology architecture to handle tough workloads at a fraction of the cost of all-flash approaches.”

The ClusterStor array line came to Seagate with its 2013, $374m Xyratex acquisition, alongside its much-wanted disk drive testing equipment. The product line includes the 9000 Big Data rack system, and the 1500 Lustre parallel file system array (including the L300 variant). There is a near-equivalent G200 Spectrum Scale parallel file system array, and an A200 active archive object store which can be a backup target for the 1500 and G200.

Within the 300N line we have the L300N and G300N products, derivatives of the L300 and G200, and intended for use in Lustre and Spectrum Scale parallel file processing environments, meaning they potentially replace or sit alongside the existing L300 Lustre and G200 Spectrum Scale systems.

Seagate tells us the 300N improves sequential performance by 33 per cent, with up to 112GB/s per rack. Also it is the highest density - 720TB, 2U - Lustre storage array, making it possible to build the world's first 15 PB 42U rack. It calls them (in a somewhat over the top pronouncement) engineered solutions.

There is a 300N embedded application controller which uses a 14nm Broadwell CPU. It supports either Intel’s 100Gbit/s Omni-Path or Mellanox InfiniBand or Ethernet links to the outside world and a 12Gbit/s internal SAS fabric, enabling up to 16GB/sec of sequential IPO per storage enclosure.

ClusterStor_300N_controller

ClusterStor 300N application controller

A ClusterStor Nytro Intelligent I/O Manager provides transparent performance acceleration for IO, checking out IO work and directing them to disk or flash as appropriate. Flash storage gets small block sizes, un-aligned I/O or read modify write workloads. The flash provides 100 per cent write acceleration and can be configured to accelerate reads as well.

Flash-held data can be automatically flushed to disk (10TB SAS drives), as SSDs fill a background task de-stages the oldest data to HDD, or, as a configuration choice, pinned in flash until an admin says otherwise.

Large data blocks are sent to the array’s so-called GridRAID disk pool.

The systems comes with a 2U, 24-drive System Management Unit (SMU) with 7 x 2.5-inch 10,000rpm drives and 5 x 2.5-inch 15,000rpm drives. There can be up to seven SSUs (Scalable Storage Units) with up to 7 in the base and expansion racks. Base rack SSUs hold 2 or 4 (1.3TB SAS or NVMe) SSDs*, and up to 82 x 3.5-inch drives spinning at either 7.200rpm or 10,000rpm; the faster speed being required for GriDRAID. Expansion rack SSUs hold up to 84 **disk drives.

PCIe x 16 NVMe flash accelerators also appear on a Seagate configuration overview slide.

ClusterStor_config_Slide

ClusterStor configuration overview slide

The L300N and G300N have Mellanox QDR/FDR/EDR InfiniBand, 40/100 GbiE*** and Intel Omni-Path network connectivity support.

The L300N has a Metadata Management Unit in addition to the System Management Unit. This holds 22 x 2.5-inch 10,000 rpm disk drives.

Mike Vildibill, VP for HPC Storage in HPE issued a supportive canned quote: “The 300N offers the density, extreme bandwidth, low latency and simplified manageability that our customers demand in their HPC storage environments today.”

HPE, which is a ClusterStor reseller is acquiring SGI which is also a ClusterStor reseller.

Seagate says burst buffers are not POSIX-compatible, may require app software changes and proprietary compute client software.

ClusterStor_300N_schedule

Other news

The A200 Active Archive, usable as a G200/G200N and L200/L200N backup device, array gets10TB drive support, meaning up to 6.5PB per rack. It also has improved orchestration facilities with ClusterStor HSM collecting data from front-end ClusterStor arrays and pumping it into the archive tier. An active archive object storage tier for G200 Spectrum Scale (GPFS) platform is coming in the next release.

At its Salt Lake City SC16 booth, number 1209, Seagate is showing an NVMe over a Fabric array using Nytro flash drives, delivering 4.8 million IOPS with single-digit microsecond latency. We understand this is a 2U enclosure with 24 Seagate Nytro XF1440 NVMe SSDs. These are front-loaded and mounted vertically in three groups of eight.

Seagate_NVMEF_Array

Seagate NVMe over fabric array with 24 x Nytro SSDs.

Mellanox ConnectX-3 or ConnectX-4 network adapters are used, and they run at speeds of 25Gb/sec or faster. Seagate also has demo HAMR (Heat Assisted Magnetic Recording) disk drives on its stand to to show its progress towards building 50TB+, 3.5-inch nearline disk drives.

Initial shipments of the ClusterStor 300N begin this quarter with general availability next quarter. The table above has more detailed system and component schedule information. ®

* Only 2 SSDs with L300N

** Only 40GbitE for the L300N

*** Only 82 with L300N expansion racks.

Similar topics


Other stories you might like

  • Lonestar plans to put datacenters in the Moon's lava tubes
    How? Founder tells The Register 'Robots… lots of robots'

    Imagine a future where racks of computer servers hum quietly in darkness below the surface of the Moon.

    Here is where some of the most important data is stored, to be left untouched for as long as can be. The idea sounds like something from science-fiction, but one startup that recently emerged from stealth is trying to turn it into a reality. Lonestar Data Holdings has a unique mission unlike any other cloud provider: to build datacenters on the Moon backing up the world's data.

    "It's inconceivable to me that we are keeping our most precious assets, our knowledge and our data, on Earth, where we're setting off bombs and burning things," Christopher Stott, founder and CEO of Lonestar, told The Register. "We need to put our assets in place off our planet, where we can keep it safe."

    Continue reading
  • Conti: Russian-backed rulers of Costa Rican hacktocracy?
    Also, Chinese IT admin jailed for deleting database, and the NSA promises no more backdoors

    In brief The notorious Russian-aligned Conti ransomware gang has upped the ante in its attack against Costa Rica, threatening to overthrow the government if it doesn't pay a $20 million ransom. 

    Costa Rican president Rodrigo Chaves said that the country is effectively at war with the gang, who in April infiltrated the government's computer systems, gaining a foothold in 27 agencies at various government levels. The US State Department has offered a $15 million reward leading to the capture of Conti's leaders, who it said have made more than $150 million from 1,000+ victims.

    Conti claimed this week that it has insiders in the Costa Rican government, the AP reported, warning that "We are determined to overthrow the government by means of a cyber attack, we have already shown you all the strength and power, you have introduced an emergency." 

    Continue reading
  • China-linked Twisted Panda caught spying on Russian defense R&D
    Because Beijing isn't above covert ops to accomplish its five-year goals

    Chinese cyberspies targeted two Russian defense institutes and possibly another research facility in Belarus, according to Check Point Research.

    The new campaign, dubbed Twisted Panda, is part of a larger, state-sponsored espionage operation that has been ongoing for several months, if not nearly a year, according to the security shop.

    In a technical analysis, the researchers detail the various malicious stages and payloads of the campaign that used sanctions-related phishing emails to attack Russian entities, which are part of the state-owned defense conglomerate Rostec Corporation.

    Continue reading
  • FTC signals crackdown on ed-tech harvesting kid's data
    Trade watchdog, and President, reminds that COPPA can ban ya

    The US Federal Trade Commission on Thursday said it intends to take action against educational technology companies that unlawfully collect data from children using online educational services.

    In a policy statement, the agency said, "Children should not have to needlessly hand over their data and forfeit their privacy in order to do their schoolwork or participate in remote learning, especially given the wide and increasing adoption of ed tech tools."

    The agency says it will scrutinize educational service providers to ensure that they are meeting their legal obligations under COPPA, the Children's Online Privacy Protection Act.

    Continue reading
  • Mysterious firm seeks to buy majority stake in Arm China
    Chinese joint venture's ousted CEO tries to hang on - who will get control?

    The saga surrounding Arm's joint venture in China just took another intriguing turn: a mysterious firm named Lotcap Group claims it has signed a letter of intent to buy a 51 percent stake in Arm China from existing investors in the country.

    In a Chinese-language press release posted Wednesday, Lotcap said it has formed a subsidiary, Lotcap Fund, to buy a majority stake in the joint venture. However, reporting by one newspaper suggested that the investment firm still needs the approval of one significant investor to gain 51 percent control of Arm China.

    The development comes a couple of weeks after Arm China said that its former CEO, Allen Wu, was refusing once again to step down from his position, despite the company's board voting in late April to replace Wu with two co-chief executives. SoftBank Group, which owns 49 percent of the Chinese venture, has been trying to unentangle Arm China from Wu as the Japanese tech investment giant plans for an initial public offering of the British parent company.

    Continue reading
  • SmartNICs power the cloud, are enterprise datacenters next?
    High pricing, lack of software make smartNICs a tough sell, despite offload potential

    SmartNICs have the potential to accelerate enterprise workloads, but don't expect to see them bring hyperscale-class efficiency to most datacenters anytime soon, ZK Research's Zeus Kerravala told The Register.

    SmartNICs are widely deployed in cloud and hyperscale datacenters as a means to offload input/output (I/O) intensive network, security, and storage operations from the CPU, freeing it up to run revenue generating tenant workloads. Some more advanced chips even offload the hypervisor to further separate the infrastructure management layer from the rest of the server.

    Despite relative success in the cloud and a flurry of innovation from the still-limited vendor SmartNIC ecosystem, including Mellanox (Nvidia), Intel, Marvell, and Xilinx (AMD), Kerravala argues that the use cases for enterprise datacenters are unlikely to resemble those of the major hyperscalers, at least in the near term.

    Continue reading

Biting the hand that feeds IT © 1998–2022