Uyuni: improve spacewalk-repo-sync performance
a project by moio
Updated almost 3 years ago. 8 hacker ♥️. 1 follower.

Let's make reposync faster

Every day,

Multiple times a day,

Every SUSE Manager customer,

Every Red Hat Satellite customer,

Every Spacewalk user,

And every Uyuni user...

...spends a lot of CPU and wall clock time in reposyncing.

Intro

A lot of that time is wasted by an old, overcomplicated and most of all inefficient algorithm that contributes heavily on heat dissipation and user patience depletion!

HackWeek hackers, we can change that!

Past attempts only partially succeeded: https://trello.com/c/inl9Wu0p/40-reduce-global-warming, https://trello.com/c/dYAR0J8K/13-reduce-global-warming-take-2

But we have better tools now!

Tooling

py-spy to the rescue: introduction

Install with: curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py python get-pip.py pip install py-spy

Trace a running spacewalk-repo-sync with: py-spy --nonblocking --pid `ps aux | grep spacewalk-repo-sync | grep -v grep | awk '{print $2}'` --flame output.svg --duration 10

Look at the results with:

python -m SimpleHTTPServer 8666

And point your browser to http://:8666/output.html. Here is one such example:

Flame Graph

Current remarks:

  • we currently spend a lot of time in lookup functions
  • lookup functions SELECT rows at every INSERT
  • this is especially bad for checksums, capabilities and some other cases
  • design comes from Oracle and can probably be changed!

Join this project

Looking for hackers with the skills:

python performance databases postgresql

This project is part of:

Hack Week 18

Activity

  • over 5 years ago: joachimwerner liked this project.
  • over 5 years ago: mkoutny liked this project.
  • over 5 years ago: Pharaoh_Atem liked this project.
  • over 5 years ago: PSuarezHernandez liked this project.
  • over 5 years ago: ebischoff liked this project.
  • over 5 years ago: moio added keyword "python" to this project.
  • over 5 years ago: moio added keyword "performance" to this project.
  • over 5 years ago: moio added keyword "databases" to this project.
  • over 5 years ago: moio added keyword "postgresql" to this project.
  • over 5 years ago: cbosdonnat liked this project.
  • over 5 years ago: moio started this project.
  • over 5 years ago: moio liked this project.
  • over 5 years ago: jbrielmaier liked this project.
  • over 5 years ago: moio originated this project.


    • Comments

    • ebischoff
      over 5 years ago by ebischoff | Reply

      See also this fate request "Have a synchronization that does not take hours (or days)"

    • joachimwerner
      over 5 years ago by joachimwerner | Reply

      Related, but probably out of scope for your hack week project: Once we've optimized the syncing code, I think we could also reduce the scope of what needs to be synced for many customers: Especially for pilots, but also in real life, many of the older updates (e.g. several complete kernels, several Java updates) are never going to be needed, but still synced. We should investigate how we can offer something like a "JeR" ("Just enough Repo") to speed things up even more. This could be done server-side (provide alternative repo metadata for a "current stuff only" repo or client-side (but then some dependency resolution magic is probably needed).

    • chasecrum
      over 5 years ago by chasecrum | Reply

      Any update on how this turned out?

    Similar Projects

    Saline (state deployment control and monitoring tool for SUSE Manager/Uyuni) by vizhestkov

    Project Description

    Saline is an addition for salt used in SUSE Manager/Uyuni aimed to provide better control and visibility for states deploymend in the large scale environments.

    In current state the published version can be used only as a Prometheus exporter and missing some of the key features implemented in PoC (not published). Now it can provide metrics related to salt events and state apply process on the minions. But there is no control on this process implemented yet.

    Continue with implementation of the missing features and improve the existing implementation:

    • authentication (need to decide how it should be/or not related to salt auth)

    • web service providing the control of states deployment

    Goal for this Hackweek

    • Implement missing key features

    • Implement the tool for state deployment control with CLI

    Resources

    https://github.com/openSUSE/saline

    Symbol Relations by hli

    Description

    There are tools to build function call graphs based on parsing source code, for example, cscope.

    This project aims to achieve a similar goal by directly parsing the disasembly (i.e. objdump) of a compiled binary. The assembly code is what the CPU sees, therefore more "direct". This may be useful in certain scenarios, such as gdb/crash debugging.

    Detailed description and Demos can be found in the README file:

    Supports x86 for now (because my customers only use x86 machines), but support for other architectures can be added easily.

    Tested with python3.6

    Goals

    Any comments are welcome.

    Resources

    https://github.com/lhb-cafe/SymbolRelations

    symrellib.py: mplements the symbol relation graph and the disassembly parser

    symrel_tracer*.py: implements tracing (-t option)

    symrel.py: "cli parser"

    SUSE AI Meets the Game Board by moio

    Use tabletopgames.ai’s open source TAG and PyTAG frameworks to apply Statistical Forward Planning and Deep Reinforcement Learning to two board games of our own design. On an all-green, all-open source, all-AWS stack!
    A chameleon playing chess in a train car, as a metaphor of SUSE AI applied to games

    Results: Infrastructure Achievements

    We successfully built and automated a containerized stack to support our AI experiments. This included:

    A screenshot of k9s and nvtop showing PyTAG running in Kubernetes with GPU acceleration

    ./deploy.sh and voilà - Kubernetes running PyTAG (k9s, above) with GPU acceleration (nvtop, below)

    Results: Game Design Insights

    Our project focused on modeling and analyzing two card games of our own design within the TAG framework:

    • Game Modeling: We implemented models for Dario's "Bamboo" and Silvio's "Totoro" and "R3" games, enabling AI agents to play thousands of games ...in minutes!
    • AI-driven optimization: By analyzing statistical data on moves, strategies, and outcomes, we iteratively tweaked the game mechanics and rules to achieve better balance and player engagement.
    • Advanced analytics: Leveraging AI agents with Monte Carlo Tree Search (MCTS) and random action selection, we compared performance metrics to identify optimal strategies and uncover opportunities for game refinement .

    Cards from the three games

    A family picture of our card games in progress. From the top: Bamboo, Totoro, R3

    Results: Learning, Collaboration, and Innovation

    Beyond technical accomplishments, the project showcased innovative approaches to coding, learning, and teamwork:

    • "Trio programming" with AI assistance: Our "trio programming" approach—two developers and GitHub Copilot—was a standout success, especially in handling slightly-repetitive but not-quite-exactly-copypaste tasks. Java as a language tends to be verbose and we found it to be fitting particularly well.
    • AI tools for reporting and documentation: We extensively used AI chatbots to streamline writing and reporting. (Including writing this report! ...but this note was added manually during edit!)
    • GPU compute expertise: Overcoming challenges with CUDA drivers and cloud infrastructure deepened our understanding of GPU-accelerated workloads in the open-source ecosystem.
    • Game design as a learning platform: By blending AI techniques with creative game design, we learned not only about AI strategies but also about making games fun, engaging, and balanced.

    Last but not least we had a lot of fun! ...and this was definitely not a chatbot generated line!

    The Context: AI + Board Games

    ClusterOps - Easily install and manage your personal kubernetes cluster by andreabenini

    Description

    ClusterOps is a Kubernetes installer and operator designed to streamline the initial configuration and ongoing maintenance of kubernetes clusters. The focus of this project is primarily on personal or local installations. However, the goal is to expand its use to encompass all installations of Kubernetes for local development purposes.
    It simplifies cluster management by automating tasks and providing just one user-friendly YAML-based configuration config.yml.

    Overview

    • Simplified Configuration: Define your desired cluster state in a simple YAML file, and ClusterOps will handle the rest.
    • Automated Setup: Automates initial cluster configuration, including network settings, storage provisioning, special requirements (for example GPUs) and essential components installation.
    • Ongoing Maintenance: Performs routine maintenance tasks such as upgrades, security updates, and resource monitoring.
    • Extensibility: Easily extend functionality with custom plugins and configurations.
    • Self-Healing: Detects and recovers from common cluster issues, ensuring stability, idempotence and reliability. Same operation can be performed multiple times without changing the result.
    • Discreet: It works only on what it knows, if you are manually configuring parts of your kubernetes and this configuration does not interfere with it you can happily continue to work on several parts and use this tool only for what is needed.

    Features

    • distribution and engine independence. Install your favorite kubernetes engine with your package manager, execute one script and you'll have a complete working environment at your disposal.
    • Basic config approach. One single config.yml file with configuration requirements (add/remove features): human readable, plain and simple. All fancy configs managed automatically (ingress, balancers, services, proxy, ...).
    • Local Builtin ContainerHub. The default installation provides a fully configured ContainerHub available locally along with the kubernetes installation. This configuration allows the user to build, upload and deploy custom container images as they were provided from external sources. Internet public sources are still available but local development can be kept in this localhost server. Builtin ClusterOps operator will be fetched from this ContainerHub registry too.
    • Kubernetes official dashboard installed as a plugin, others planned too (k9s for example).
    • Kubevirt plugin installed and properly configured. Unleash the power of classic virtualization (KVM+QEMU) on top of Kubernetes and manage your entire system from there, libvirtd and virsh libs are required.
    • One operator to rule them all. The installation script configures your machine automatically during installation and adds one kubernetes operator to manage your local cluster. From there the operator takes care of the cluster on your behalf.
    • Clean installation and removal. Just test it, when you are done just use the same program to uninstall everything without leaving configs (or pods) behind.

    Planned features (Wishlist / TODOs)

    • Containerized Data Importer (CDI). Persistent storage management add-on for Kubernetes to provide a declarative way of building and importing Virtual Machine Disks on PVCs for

    Run local LLMs with Ollama and explore possible integrations with Uyuni by PSuarezHernandez

    Description

    Using Ollama you can easily run different LLM models in your local computer. This project is about exploring Ollama, testing different LLMs and try to fine tune them. Also, explore potential ways of integration with Uyuni.

    Goals

    • Explore Ollama
    • Test different models
    • Fine tuning
    • Explore possible integration in Uyuni

    Resources

    • https://ollama.com/
    • https://huggingface.co/
    • https://apeatling.com/articles/part-2-building-your-training-data-for-fine-tuning/