mlerner

6 ms·

1. 1.▲
   Graham: Synchronizing Clocks by Leveraging Local Clock Properties (2022) [pdf](usenix.org)

   59 points by mlerner 1y ago|13 comments
2. 2.▲
   Autothrottle: Resource Management for SLO-Targeted Microservices(usenix.org)

   22 points by mlerner 2y ago|1 comments
3. 3.▲
   Resiliency at Scale: Managing Google's TPUv4 Machine Learning Supercomputer(micahlerner.com)

   1 points by mlerner 2y ago|0 comments
4. 4.▲
   How the anti-cheat is anti-cheating so far(leagueoflegends.com)

   3 points by mlerner 2y ago|0 comments
5. 5.▲
   ServiceRouter: Hyperscale and Minimal Cost Service Mesh at Meta(micahlerner.com)

   1 points by mlerner 3y ago|0 comments
6. 6.▲
   A Cloud-Scale Characterization of Remote Procedure Calls(newsletter.micahlerner.com)

   2 points by mlerner 3y ago|0 comments
7. 7.▲
   A Cloud-Scale Characterization of Google's Remote Procedure Calls(micahlerner.com)

   1 points by mlerner 3y ago|0 comments
8. 8.▲
   Gemini, Amazon's system for fast failure recovery in distributed model training(micahlerner.com)

   2 points by mlerner 3y ago|0 comments
9. 9.▲
   Defcon: Preventing overload with graceful feature degradation (2023)(micahlerner.com)

   237 points by mlerner 3y ago|95 comments
10. 10.▲
    MotherDuck: DuckDB in the Cloud and in the Client [pdf](cidrdb.org)

    5 points by mlerner 3y ago|0 comments
11. 11.▲
    Gemini: Fast Failure Recovery in Distributed Training with In-Memory Checkpoints(micahlerner.com)

    3 points by mlerner 3y ago|0 comments
12. 12.▲
    by mlerner 3y ago

    Author of the paper summary here - my understanding is that XFaaS doesn't run functions that are run in response to user input (e.g. XFaaS does not execute code that fetches and returns data because a user clicked on a button).
13. 13.▲
    Gemini: Fast Failure Recovery in Distributed Training with In-Memory Checkpoints(newsletter.micahlerner.com)

    4 points by mlerner 3y ago|0 comments
14. 14.▲
    XFaaS: Hyperscale and Low Cost Serverless Functions at Meta(newsletter.micahlerner.com)

    3 points by mlerner 3y ago|0 comments
15. 15.▲
    XFaaS: Hyperscale and Low Cost Serverless Functions at Meta(newsletter.micahlerner.com)

    3 points by mlerner 3y ago|1 comments
16. 16.▲
    Blueprint: A Toolchain for Highly-Reconfigurable Microservice Applications(micahlerner.com)

    2 points by mlerner 3y ago|0 comments
17. 17.▲
    Gemini: Fast Failure Recovery in Distributed Training with In-Memory Checkpoints [pdf](cs.rice.edu)

    50 points by mlerner 3y ago|13 comments
18. 18.▲
    XFaaS: Hyperscale and Low Cost Serverless Functions at Meta [pdf](cis.upenn.edu)

    4 points by mlerner 3y ago|0 comments
19. 19.▲
    Efficient Memory Management for Large Language Model Serving with PagedAttention(newsletter.micahlerner.com)

    3 points by mlerner 3y ago|0 comments
20. 20.▲
    Efficient Memory Management for Large Language Model Serving with PagedAttention(newsletter.micahlerner.com)

    1 points by mlerner 3y ago|0 comments
21. 21.▲
    Blueprint: A Toolchain for Highly-Reconfigurable Microservice Applications(micahlerner.com)

    1 points by mlerner 3y ago|0 comments
22. 22.▲
    by mlerner 3y ago

    Good catch! Here's the link to the framework: https://gitlab.mpi-sws.org/cld/blueprint/blueprint-compiler . Also added it to the blog post :)
23. 23.▲
    Blueprint: A Toolchain for Highly-Reconfigurable Microservice Applications(micahlerner.com)

    5 points by mlerner 3y ago|2 comments
24. 24.▲
    Defcon: Preventing Overload with Graceful Feature Degradation(micahlerner.com)

    4 points by mlerner 3y ago|1 comments
25. 25.▲
    Defcon: Preventing Overload with Graceful Feature Degradation(micahlerner.com)

    4 points by mlerner 3y ago|0 comments
26. 26.▲
    by mlerner 3y ago

    The x-axis is size of trade, and the y-axis is % of trades at a specific size. ExRates had _many_ large trades, which shows up as a significantly different distribution than other exchanges like Coinbase/Bitstamp/Kraken (which had
27. 27.▲
    by mlerner 3y ago

    100% - although this time around there is (in my opinion) _a lot more data_ [1] that regulated markets (like CME bitcoin futures) drive price discovery. [1] There was data back then too, it just wasn't received.
28. 28.▲
    by mlerner 3y ago

    The volume we were looking at was on exchanges (not on chain).
29. 29.▲
    by mlerner 3y ago

    It’s Jekyll - and open source! https://github.com/mlerner/mlerner.github.io
30. 30.▲
    by mlerner 3y ago

    Neat and funny that they have an uncited figure (the one with the six exchanges with wacky trade size distributions) from research that we did while I was working at Bitwise. We found 95% of bitcoin trading volume at the time was fake: htt