Walk with me... through a billion files. Slow down – admire the subset

Qumulo and the tree-walking problem


Analysis If you ask your notebook's filesystem how many MP3 files it is storing that haven’t been opened in 30 days, you can find the answer reasonably quickly. But ask an enterprise’s file system when it holds a million files and you have a big problem.

Ask this question of a file system that holds a billion files and your day just got a whole lot worse.

Here's a filesystem 101 to say why this happens.

A file system is like an upside-down tree structure of files in folders, with the folders forming a directory tree descending from a single root. Each node in this structure lists the files it contains, plus data about the files, and the sub-folders in contains. There is no central directory in one place listing all this stuff for the entire filesystem.

Filesystem_tree

Upside-down file system tree structure

So, to answer the initial question: the system has to traverse or walk the file system tree and at each node (F1-1, F1-2) look for files with the .MP3 extension and their last opened date, adding them to a list if they match the filter criteria. If there is a nested sub-folder or sub-folders (F2-1, F2-2) it has to walk the tree down to the first one ( F2-1) node and repeat the process, and then repeat again for any sub-folder, until it gets to the bottom of that series of nodes, then walk back up until it gets to a node where there is another sub-folder (F2-2) listed and go to that, and so on ad infinitum, meaning the end of the file system.

Assume each node access requires a disk access and this takes 10 milliseconds; then a 10-node file system would take 100ms roughly plus the access needed to walk back up the tree; say 150ms being simplistic.

So, again, a 100-node file system would take 1,500ms, a thousand node one 15,000ms, a million node one 15,000,000ms and a billion node one 15,000,000,000ms - like we said, your day just got a whole lot worse because the tree walk is going to take days, 173.6 if our often suspect math is correct.

Qumulo CTO and co-founder Peter Godman, presenting to a press briefing, says these kind of numbers aren’t imaginary. A major DreamWorks picture needs 500 million files and legacy kit – meaning pre-2010 – can’t cope with this kind of filesystem request, going into a kind of tree-walk paralysis, which makes them hard to manage and optimise.

Qumulo says tree walks make data management tasks days to weeks long, leading to data blindness.

Peter_Goodman

Qumulo co-founder and CEO Peter Godman

Seemingly simple requests – such as how many MP3 files there are that haven’t been opened in 30 days – are practically impossible to accomplish, let alone telling the filesystem to move them off to cheap back-end cloud storage.

Qumulo’s marketing VP, Jay Wampold, says: “When you have a billion anything, humans can’t manage it.”

As Godman says, the metadata processing involved becomes a problem in its own right: ”The metadata itself is a big data problem at scale” and with QF2 (Qumulo File Fabric) we have real-time control of files at scale.

You can’t retrofit the necessary metadata generation, storage and access to an existing file system. It has to be designed in, which it has been when it comes to Qumulo’s scale-out filesystem (QFS, with its underlying scalable block store (SBS).

There is a Qumulo database component, an extension of traditional file system metadata, which puts virtual fields in file metadata, and has an analytics capability.

Qumulo invented it and built it and it is distributed across nodes. It is the firm's own metadata database, and a property of its file system tree, not a separate "box" containing metadata.

A Qumulo QF2 technical overview (PDF) declares:

When you have a large numbers of files, the directory structure and file attributes themselves become big data. As a result, sequential processes such as tree walks, which are fundamental to legacy storage, are no longer computationally feasible. Instead, querying a large file system and managing it requires a new approach that uses parallel and distributed algorithms.

The technical paper says: “There is an inode B-tree, which acts as an index of all the files. The inode list is a standard file-system implementation technique that makes checking the consistency of the file system independent of the directory hierarchy. Inodes also help to make update operations such as directory moves efficient.

Files and directories are represented as B-trees with their own key/value pairs, such as the file name, its size and its access control list (ACL) or POSIX permissions.

This reliance on B-trees that point to virtualized protected block storage in SBS is one of the reasons that in QF2, a file system with a trillion files is feasible.

QumuloDB analytics are built in and integrated with the file system itself. Because the QF2 file system relies on B-trees, the analytics can use a system of real-time aggregates and information is available for timely processing without costly file system tree walks.

Read the technical overview to find out more. ®

Similar topics


Other stories you might like

  • Talos names eight deadly sins in widely used industrial software
    Entire swaths of gear relies on vulnerability-laden Open Automation Software (OAS)

    A researcher at Cisco's Talos threat intelligence team found eight vulnerabilities in the Open Automation Software (OAS) platform that, if exploited, could enable a bad actor to access a device and run code on a targeted system.

    The OAS platform is widely used by a range of industrial enterprises, essentially facilitating the transfer of data within an IT environment between hardware and software and playing a central role in organizations' industrial Internet of Things (IIoT) efforts. It touches a range of devices, including PLCs and OPCs and IoT devices, as well as custom applications and APIs, databases and edge systems.

    Companies like Volvo, General Dynamics, JBT Aerotech and wind-turbine maker AES are among the users of the OAS platform.

    Continue reading
  • Despite global uncertainty, $500m hit doesn't rattle Nvidia execs
    CEO acknowledges impact of war, pandemic but says fundamentals ‘are really good’

    Nvidia is expecting a $500 million hit to its global datacenter and consumer business in the second quarter due to COVID lockdowns in China and Russia's invasion of Ukraine. Despite those and other macroeconomic concerns, executives are still optimistic about future prospects.

    "The full impact and duration of the war in Ukraine and COVID lockdowns in China is difficult to predict. However, the impact of our technology and our market opportunities remain unchanged," said Jensen Huang, Nvidia's CEO and co-founder, during the company's first-quarter earnings call.

    Those two statements might sound a little contradictory, including to some investors, particularly following the stock selloff yesterday after concerns over Russia and China prompted Nvidia to issue lower-than-expected guidance for second-quarter revenue.

    Continue reading
  • Another AI supercomputer from HPE: Champollion lands in France
    That's the second in a week following similar system in Munich also aimed at researchers

    HPE is lifting the lid on a new AI supercomputer – the second this week – aimed at building and training larger machine learning models to underpin research.

    Based at HPE's Center of Excellence in Grenoble, France, the new supercomputer is to be named Champollion after the French scholar who made advances in deciphering Egyptian hieroglyphs in the 19th century. It was built in partnership with Nvidia using AMD-based Apollo computer nodes fitted with Nvidia's A100 GPUs.

    Champollion brings together HPC and purpose-built AI technologies to train machine learning models at scale and unlock results faster, HPE said. HPE already provides HPC and AI resources from its Grenoble facilities for customers, and the broader research community to access, and said it plans to provide access to Champollion for scientists and engineers globally to accelerate testing of their AI models and research.

    Continue reading

Biting the hand that feeds IT © 1998–2022