home.social

#faulttolerance — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #faulttolerance, aggregated by home.social.

  1. If the Transaction Coordinator crashes after collecting votes but before sending its decision, participants remain blocked until recovery. This blocking behavior is a known limitation of 2PC in distributed fintech systems.

    #DistributedSystems #Fintech #FaultTolerance

  2. If a lock holder crashes, the lock must expire via time-to-live. If the TTL is too short, a slow process may lose its lock while still executing, allowing a second process to proceed — a correctness failure in fintech.

    #DistributedLocking #FaultTolerance #Fintech

  3. If a lock holder crashes, the lock must expire via time-to-live. If the TTL is too short, a slow process may lose its lock while still executing, allowing a second process to proceed — a correctness failure in fintech.

    #DistributedLocking #FaultTolerance #Fintech

  4. Consensus requires a quorum — a majority of nodes — to agree before a value is committed. This majority requirement is what allows the system to continue despite node failures.

    #Consensus #FaultTolerance #Fintech

  5. So apparently, NASA's secret sauce for building a "fault-tolerant" computer involves getting blocked by #Cloudflare while trying to access the article. 🚫✨ Who knew #cybersecurity was just a fancy way of saying you can't read about computers? 🤖🔒
    cacm.acm.org/news/how-nasa-bui #NASA #FaultTolerance #ComputerScience #HackerNews #ngated

  6. For those with more than a passing interest in Information Systems security (OK not everyone🤣) , this might prove to be interesting. Although as commented,

    ’build the governance first, get the key parties committed, define the trust roots, enforce the rules – is precisely the kind of process that works in Switzerland and struggles almost everywhere else.’

    theregister.com/2026/03/17/swi

    #IT #Security #FaultTolerance #ETH #Switzerland

  7. #Uber redesigned its #MySQL fleet using a consensus-driven architecture based on MySQL Group Replication:
    ✅ Cluster failover dropped from minutes → seconds
    ✅ Leader election & failure detection now inside the database layer
    ✅ Improved availability, simpler orchestration, stronger consistency across thousands of production clusters

    Learn more: bit.ly/4b8RWYL

    #SoftwareArchitecture #DistributedSystems #Clusters #FaultTolerance #RelationalDatabases

  8. 🔥 Behold the #PyTorch blog masterpiece: "Fault Tolerant #Llama Training" - because who doesn't love 2000 failures every 15 seconds? 😂💥 Forget checkpoints, because llamas are clearly bred for #chaos on a Crusoe L40S! 🙄✨
    pytorch.org/blog/fault-toleran #Training #FaultTolerance #MachineLearning #HackerNews #ngated

  9. Reviving the Legacy of Tandem Computers: Pioneers of Fault-Tolerant Technology

    Once a titan in the realm of fault-tolerant computing, Tandem Computers revolutionized transaction processing systems. As we delve into its innovative architecture and enduring impact on modern comput...

    news.lavx.hu/article/reviving-

    #news #tech #FaultTolerance #NonStopComputing #TransactionProcessing

  10. #CellBasedArchitecture is revolutionizing the way we build resilient systems. This architecture enables each cell to manage its resources and make decisions autonomously by emphasizing core principles such as isolation, autonomy & replication.

    #Observability for cell-based architecture requires a tailored approach to address the unique challenges and opportunities presented by this distributed system design.

    #InfoQ article by Yury Niño Roa: bit.ly/40i72G6

    #DistributedSystems #Resilience #Microservices #FaultTolerance

  11. I am a #CitizenScientist at heart, and it bothers me not that #technology is prevalent and powerful (high speed, high storage capacity), but that such improvement of our technology is gatekept behind massive #paywalls.

    (TL;DR: the cost of a portable genome sequencer and also data storage makes it unattainable to collaborate as a citizen scientist, and that makes me big mad 😡!! I know we can do something about it.)

    I dream of an internet #infrastructure that would enable me, and others, to #democratize data, storage space, and processing power. Such would bolster research and study across many fields of knowledge.

    Allow me to set the scene.

    - A #PublicCommons #CommunityMesh #MeshNetwork interlinks public and private schools, weather broadcast stations, public government buildings, medical facilities, public transit centers, large housing complexes, and observation towers in public parks for civil and public academic use. Low-power Radio Wide Area Networks (#LoRaWAN) might prove useful here for environmental sensors.

    - #FreedomBox or no-assembly-required #SelfHosting #cloudlet solutions bring #FaultTolerance to the hands of #WebService architects.

    - #NetworkAttachedStorage (#NAS) equips locations connected to the network with an opportunity to share _some_ data storage capacity as a cache for the functions of the greater network. (In my opinion, altruistic features should be opt-out conditions.) Such might confer a cache of operable @wikifunctions representations.

    - #UniformResourceName (#URN) resolvers (e.g., such as for #DOI, #ARC, #ISBN, #IPFS) give us what we need to collaborate as communities to organize our #DataSets and to distribute amongst our community the load of resolving names of data to their resources.

    - the #ResourceDescriptionFramework, or #RDF, lets us claim to properties, attributes and metainformation of a resource. This is useful for cataloging and searching through datasets to find pertinent information. (E.g., by category: music, video, lectures; photos, figures, charts, diagrams; databases, manifests, ledgers, logs; etc.)

    - #Peer2Peer clients create a distributed #ComputeFabric for domains of interest.

    Major hubs of the network publish off-loadable functions in "initiatives" to: collect, clean, stage, process, recollect, and report sets of data for research within their pertinent areas of interest. (E.g., universities or parks might produce functions to operationalize a project to collect and study evidence of climate change.)

    Client-Devices subscribe to the initiatives that their owner finds worthwhile and commit to serving requests they receive for such functions.

    --------

    The scene is set. Now we imagine:

    You subscribe to a citizen science initiative and travel anywhere with your portable sensor(s): a #GenomeSequencer or #Spectrometer, a #SchumannResonance or #Radiation detector, a #Barometer or #Thermometer or #Hygrometer or #Decibellometer.

    You take #measurements, name them and fill in whatever #metadata about them you can. When you submit your dataset to the network, the publishers of every #initiative of which you are part are notified so that others can search for it through them via the metadata you ascribed to the dataset. When others search for and find your dataset, they, too, can process it according the functions within the network.

    By this our internet would become a virtual meta-machine; this could assist in the symbiotization of #civilian #academy and #inventive #altruism.

    (TL;DR: the cost of portable genome sequencers and data storage makes it hard to collaborate as a citizen scientist, and that makes me big mad 😡!! I know we can do something about it.)

  12. I am a #CitizenScientist at heart, and it bothers me not that #technology is prevalent and powerful (high speed, high storage capacity), but that such improvement of our technology is gatekept behind massive #paywalls.

    (TL;DR: the cost of a portable genome sequencer and also data storage makes it unattainable to collaborate as a citizen scientist, and that makes me big mad 😡!! I know we can do something about it.)

    I dream of an internet #infrastructure that would enable me, and others, to #democratize data, storage space, and processing power. Such would bolster research and study across many fields of knowledge.

    Allow me to set the scene.

    - A #PublicCommons #CommunityMesh #MeshNetwork interlinks public and private schools, weather broadcast stations, public government buildings, medical facilities, public transit centers, large housing complexes, and observation towers in public parks for civil and public academic use. Low-power Radio Wide Area Networks (#LoRaWAN) might prove useful here for environmental sensors.

    - #FreedomBox or no-assembly-required #SelfHosting #cloudlet solutions bring #FaultTolerance to the hands of #WebService architects.

    - #NetworkAttachedStorage (#NAS) equips locations connected to the network with an opportunity to share _some_ data storage capacity as a cache for the functions of the greater network. (In my opinion, altruistic features should be opt-out conditions.) Such might confer a cache of operable @wikifunctions representations.

    - #UniformResourceName (#URN) resolvers (e.g., such as for #DOI, #ARC, #ISBN, #IPFS) give us what we need to collaborate as communities to organize our #DataSets and to distribute amongst our community the load of resolving names of data to their resources.

    - the #ResourceDescriptionFramework, or #RDF, lets us claim to properties, attributes and metainformation of a resource. This is useful for cataloging and searching through datasets to find pertinent information. (E.g., by category: music, video, lectures; photos, figures, charts, diagrams; databases, manifests, ledgers, logs; etc.)

    - #Peer2Peer clients create a distributed #ComputeFabric for domains of interest.

    Major hubs of the network publish off-loadable functions in "initiatives" to: collect, clean, stage, process, recollect, and report sets of data for research within their pertinent areas of interest. (E.g., universities or parks might produce functions to operationalize a project to collect and study evidence of climate change.)

    Client-Devices subscribe to the initiatives that their owner finds worthwhile and commit to serving requests they receive for such functions.

    --------

    The scene is set. Now we imagine:

    You subscribe to a citizen science initiative and travel anywhere with your portable sensor(s): a #GenomeSequencer or #Spectrometer, a #SchumannResonance or #Radiation detector, a #Barometer or #Thermometer or #Hygrometer or #Decibellometer.

    You take #measurements, name them and fill in whatever #metadata about them you can. When you submit your dataset to the network, the publishers of every #initiative of which you are part are notified so that others can search for it through them via the metadata you ascribed to the dataset. When others search for and find your dataset, they, too, can process it according the functions within the network.

    By this our internet would become a virtual meta-machine; this could assist in the symbiotization of #civilian #academy and #inventive #altruism.

    (TL;DR: the cost of portable genome sequencers and data storage makes it hard to collaborate as a citizen scientist, and that makes me big mad 😡!! I know we can do something about it.)

  13. #Google's secret to a reliable Spanner database is #ChaosTesting!

    Find out how they use it to inject faults into production-like instances and stress the system's ability to behave correctly in the face of unexpected failures.

    Explore the details on #InfoQ: bit.ly/3yyhMV8

    #ChaosEngineering #DevOps #Reliability #FaultTolerance #Testing

  14. 🚀 What is your knowledge about High Availability (HA) software? Learn all about its key features and functions in our latest Knowledge Base article! From fault tolerance to automated failover, ensures continuous operation and accessibility with minimal downtime. Check it out!

    relianoid.com/resources/knowle

  15. 🚀 Unlocking Peak Performance: Adaptive Load Balancing in Distributed Computing Discover how dynamically optimizes resource allocation, driving and in modern architectures. Explore its role in achieving and cost efficiency, ensuring resilient and responsive services.

    relianoid.com/resources/knowle

  16. I get a bit frustrated sometimes when really smart people say they don't use microservices simply because their applications "...don't need that kind of scalability.". While that is probably true, there are MANY other reasons microservices might be appropriate. Reliability & fault-tolerance as well as developer satisfaction spring to mind.

    cognitiveinheritance.com