Sharing practical vivo Internet technology insights and salon events, plus the latest industry news and hot conferences.
The article examines a real‑world memory alarm in a Vivo service, explains how glibc’s ptmalloc allocator manages heap memory via brk, sbrk, and mmap, shows why freed chunks stay in bins, and recommends limiting heap growth or switching to jemalloc for faster reclamation.
At the 19th China Linux Kernel Developer Conference, vivo unveiled nine open‑source innovations—including DMA‑BUF DIO for AI model loading, ZHP memory compression, EROFS and f2fs filesystem enhancements, and I/O optimizations—contributing significant performance gains and demonstrating its commitment to Linux kernel advancement worldwide.
This guide explains how to diagnose and fix TiKV disk‑space problems by identifying oversized log files, redundant space‑placeholder files, and excessive RocksDB/Titan data, offering command‑line checks, configuration tweaks such as enabling log rotation, disabling reserve space, and tuning GC and Titan discardable‑ratio to restore balanced storage.
Technical debt, introduced by Ward Cunningham, describes the hidden cost of rushed or sub‑optimal code that degrades efficiency, quality, and developer experience; it arises from intentional or unintentional shortcuts across four dimensions, and can be managed through systematic identification, analysis, resolution, review, and dedicated ownership, yielding more stable, higher‑quality software.
This article presents a Sentinel‑based dynamic rate‑limiting solution that uses parameter priority and a sliding‑window counter to automatically adjust thresholds per scenario, addressing the limitations of static QPS limits, simplifying rule management via Zookeeper, and integrating as an independent SDK within Sentinel’s responsibility‑chain architecture.
To simplify large‑scale Vue upgrades such as Vue 2→Vue 3, Webpack 4→Webpack 5, or migration to Vite, the author presents an automated AST‑based migration scaffold that parses JavaScript, Vue, and style files, applies transformation rules with GoGoCode and PostCSS, and rewrites code, reducing manual effort and error risk.
Vivo’s full‑link multi‑version deployment framework lets a complete microservice system be launched once while concurrently running multiple service versions in isolated baseline and on‑demand feature environments, using custom traffic‑coloring headers, tag propagation across HTTP/Dubbo/MQ, dynamic routing, and distributed tracing to reduce resource waste, accelerate hot‑fix testing, and pave the way for future data‑level isolation and production‑grade A/B experiments.
Vivo’s storage team redesigned Redis cluster slot migration by disguising the target as a slave and using master‑slave replication with progressive loading, which eliminates ask‑move round trips, reduces latency spikes, consolidates topology updates, and ensures stable access and high availability during horizontal scaling.
The article examines a single sign‑on loop‑login problem caused by Secure‑flagged cookies being set over HTTP, explains how missing tokens trigger endless redirects, and recommends enforcing HTTPS or using a non‑Secure auxiliary token to break the redirect cycle.
The article chronicles the time wheel algorithm from its 1997 origins, explains simple, round‑based, and hierarchical wheel designs, compares them to queue timers, details Dubbo’s HashedWheelTimer implementation, and shows how combining the wheel with a delay‑message queue can dramatically improve service scalability and efficiency.
