Apache Arrow的动态

Hi #All ?? Snowflake ?? #DataJedis & #DataSaiyans! ?????New feature is coming! ?? Server less tasks + [Scala, Python, Java] code process ?? Yes, You can NOW! 1?? On one side Serverless Tasks: Tasks that rely on Snowflake to determine the ideal size of compute resources for a given run based on a dynamic analysis of statistics for the most recent runs of the same task. 2?? Snowflake suplorts a wide broadband of programming language codes (Java, Scala, Python and SQL) in order to ease your migration and compatibility, and facilitate your technology stack adaptability. ???? Do you imagine what you can do now with these new capabilities in your Datapipelines? ?? Let your imagination fly, the sky is your limit, the technology and use cases are not. Release note: https://lnkd.in/db3QVRzG Supported languages: https://lnkd.in/dTqpw7pJ May the ?? be with you! #Snowflake #ServerlessTasks #dataworkload #ProgrammingCode #snowflakedevelopers #SnowflakeSquad #Snowflake_squad #Python #Java #SQL #Scala

Today's SO - an ugly hack that is going to silently break prod someday. https://lnkd.in/eBDEZmup People do this all the time. This is really an X/Y problem. "I just need..." X/Y Problem explained: https://lnkd.in/eaEMc_e7 What you really need is to get your workplace to see the risk and let you install the tool that mitigates it properly... But yes, sometimes stupid policies prevent the *right* way from being an option, so you have to hack together what you can. Don't let "perfect" be the enemy of "good". I wrote an article a while back on this, too, if you are interested - https://lnkd.in/e24cbF5G Use the right tools if it's AT ALL possible, though.? Otherwise, get used to those 2am outage calls on Sunday mornings.

Raft is a distributed consensus algorithm. Here is a simple way to understand what problem such algorithms and Raft in particular solves: Since it’s already summer, imagine there is an ice-cream delivery server which manages order info of various customers. There are many reasons the ice-cream company may want to run multiple instances of this server, one of them being availability i.e. even if one crashes, you can still order your ice-cream. But once we have multiple instances, consistency can take a hit. What if when your ice-cream is about to be dispatched, one server says you asked for mango flavor, but another says you asked for orange! (Maybe in reality, you asked for orange, but later changed to mango and the second server is just lagging) This is what distributed (i.e. multiple ice cream servers) consensus (i.e agree on the same flavor) is. One popular way to solve this problem is with a leader-follower server approach i.e. for consistency (strong, not eventual) one can always query the leader which is supposed to have the most updated state. In the background, the leader replicates its state to all the followers, so that one of them can become leader in case the current leader crashes. Raft, among other things, provides a leader election algorithm. Interesting thing about Raft is that it was designed for “understandability” as a metric. The consensus algorithm before Raft, “Paxos” , was notoriously difficult to understand, limiting the number of its practical implementations and rate of improvement. I am trying to understand the Raft algorithm myself, and wouldn’t attempt simplifying its details just yet. In case this sounds interesting to you, you may want to check out the original Raft paper linked in comments! #programming?#softwareengineering?#computerscience?#computernetworks?#Database?#DistributedSystems #Raft

?? K4 v1.5.1 out now!! ?? Now offering background flushing and compaction for further optimized read and writes. https://lnkd.in/gQNKu42X #database #storageengine #opensource #lsmtree #k4 #golang #keyvalue #algorithms

??? Calculate the cost of self-support taking into account hidden variables. Read more in this blog post: https://bit.ly/4e3yUlR

It might be helpful in some cases. https://lnkd.in/eVfNQWcR

Hey all, this is my very first post of my blog. On this post I'm talking about a framework called Carter, which is intended to an extension of Minimal API's. Hope you guys enjoy. https://lnkd.in/gq9Vs79e

?? First release of 2025! #osctrl 0.4.2 is available! This release brings significant improvements to our open-source osquery management solution. Around 5 five times less backend usage! ?? ?? Security Enhancements: * Updated x/crypto library to address ??????-????????-?????????? * Enhanced token expiration handling in ????????????-?????????? and ????????????-?????? ??? Core Improvements: * Major refactoring of distributed query system for better performance * Added support for #osquery 5.14.1 * Implemented configurable log levels and formats across services * Fixed critical issues preventing query completion * Improved elastic logger functionality for ????????????-?????? ?? System Optimization: * Cleaned up log caching mechanism * Removed legacy ingested metrics * Enhanced startup messages and logging configurations A huge thank you to all contributors who made this release possible ?? ?? See all changes: https://lnkd.in/dXiHU7wY #detection #Infrastructure #Security #DFIR #Compliance #IT #DevOps #security

Complete Guide: SAML 2.0 Assertion Grant Type (urn:ietf:params:oauth:grant-type:saml2-bearer) explained with a use case https://lnkd.in/gKHF8Pm3

See #flowrun in action! Really useful for explaining algorithms and flows.