网游服务器 Services-based 和 Cells-based 架构

本文讨论了两种网游服务器架构——基于服务的架构与基于cell的架构,并提出无缝大地图网游应采用基于cell的架构实现地图功能,其他功能则按服务实现。实时性强的功能如PK应在cell上实现以减小延迟。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

网游服务器 Services-based 和 Cells-based 架构

(金庆的专栏 2018.3)

无缝世界网游服务器架构的设计思路
http://blog.youkuaiyun.com/SmartTony/article/details/6842065
提出了2种网游服务器架构:

  • Services-based architecture 基于服务的架构
  • Cells-based architecture 基于cell的架构

基于服务的架构按功能划分成不同服务,不同服务实现于不同的进程。
而基于cell的架构使用相同的Cell进程组成一个集群,每个Cell进程包含全部的功能。

基于服务的架构一直是主流的架构,自从 microservice 和 service mesh 概念相继流行,
基于微服务的架构应该会成为未来的方向。

在无缝大地图网游中,应该用基于cell的架构来实现地图及相关功能,
其他与位置无关的功能则应该尽量按服务实现。
考虑到服务来回调用延时较大,对于实时性强的功能,如PK也应该在cell上实现。

可以考虑客户端通过专门的网关来连接cell, 使用udp/kcp来减小时延。
而其他非实时性服务由Tcp网关转发。
服务器内部可全部统一采用 gRPC.

### Short Read mNGS Technology Overview Short read metagenomics next generation sequencing (mNGS) has become a powerful tool due to its ability to provide rapid and cost-effective insights into complex microbial communities. When NGS technologies emerged, these methods significantly reduced costs while increasing speed for genome sequencing projects[^1]. This advancement allowed researchers to explore diverse microbiomes at unprecedented resolution. In short read mNGS workflows, DNA is fragmented into smaller pieces suitable for high-throughput sequencers like Illumina platforms. These fragments undergo amplification before being sequenced in parallel across millions of clusters on flow cells. The resulting sequences typically range from 100-300 base pairs long, offering sufficient coverage depth but limited contiguity compared to longer reads[^2]. Despite their shorter length, short reads offer several advantages: - **High throughput**: Capable of generating vast amounts of data quickly. - **Cost efficiency**: Lower per-basepair sequencing expenses make large-scale studies feasible. - **Wide adoption**: Established bioinformatics pipelines support robust downstream analyses. However, challenges remain regarding assembly accuracy and resolving highly repetitive or polymorphic regions within genomes. For certain applications requiring higher fidelity reconstructions—such as de novo assemblies or detailed structural variant detection—the limitations associated with short reads may necessitate complementary approaches using long-read technologies. ### Applications of Short Read mNGS #### Microbial Community Profiling One primary application involves characterizing the composition and diversity of microorganisms present in various environments such as soil, water bodies, human gut flora, etc. By mapping generated sequence reads against reference databases, taxonomic classification can be achieved down to species level often revealing novel organisms previously unknown. #### Pathogen Detection & Surveillance Clinical diagnostics benefit greatly through direct pathogen identification directly from patient samples without prior culture enrichment steps. Rapid turnaround times facilitate timely intervention strategies during outbreaks or infections where conventional methods might delay critical decisions. #### Functional Metagenomics Studies Beyond mere presence-absence calls, functional annotations derived via gene prediction algorithms enable understanding metabolic pathways active among community members under investigation. Insights gained help elucidate ecological roles played by different taxa contributing towards ecosystem services provided. ```python # Example Python code snippet demonstrating basic workflow step involving quality control filtering raw fastq files from Bio import SeqIO def filter_reads(input_file, output_file, min_length=75): filtered_records = [] # Parse input FASTQ file records = list(SeqIO.parse(input_file, "fastq")) # Filter based on minimum acceptable read length for record in records: if len(record.seq) >= min_length: filtered_records.append(record) # Write qualified reads back out to new FASTQ file SeqIO.write(filtered_records, output_file, "fastq") filter_reads('raw_data.fastq', 'filtered_data.fastq') ```
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值