跟讀練習: How Uber Scaled from One Server to 150 Million Reads Per Second - 通過影片學習英語口語

選擇練習模式

每種模式都會開啟全屏練習室。

134 句
如果句子過短或過長,請點選 Edit 進行調整。
1
In 2009, Uber was starting with a Python mullet.
2
One application, one growing business.
3
By 2019, Uber reported more than 15 million trips a day.
4
And in 2025, its caching infrastructure reported over 150 million cache database row reads per second at peak.
5
So how does a system grow at that scale?
6
Let's build it step by step.
7
Imagine you are building Uber you start with one application
8
and a postgres database and this is called monolith
9
and all your main application features live inside one application
10
and early days that's not a bad thing it's easier to build test
11
and deploy code but now more rides are booing trips more drivers are sending updates
12
and your application has work to do if the application and database share one machine they compete for the CPU and RAM.
13
The database run queries, the application handle request and both need resources.
14
One solution is to move the database onto its own server.
15
Now the application and database have separate resources.
16
You can scale them independently but there is a catch.
17
They now communicate over a network and the database can still become a bottleneck.
18
Meanwhile your application is also getting more traffic.
19
So let's scale your application first the simplest option is to give your application server more cpu
20
and memory and this is called vertical scaling it can work well
21
when you are just starting out you don't need to redesign everything just
22
because traffic increases but one machine has limits and
23
if it fails your application may become unavailable
24
so instead of making one server bigger you can add more servers
25
and that is horizontal scaling now you have multiple servers handling request
26
but there's a new question when the request arrives
27
which server should handle it
28
so you put a load of answer in front of replication server
29
it distributes incoming requests across them it can use strategies like round robin
30
or lease connection
31
and health checks can help it avoid sending traffic to unhealthy server
32
now one server failing does not necessarily take down the whole application
33
but there is a other problem imagine a user logs in replication store their session on the first server
34
then the user's next request reaches a different server
35
that server may not have the session so how does it know
36
that you are logged in one option is to store session in a shared system such as redis
37
now any application server can retrieve the session
38
and other option is to use a token such as gwd
39
that carries signed claims that can reduce the need to look up a session on every request.
40
But remember, tokens don't eliminate every kind of server -side state.
41
The key idea is simple.
42
Don't make one application server the only place where information needed by the next request exists.
43
We've added application servers, but all of them still depend on database.
44
And every request does not necessarily need to read fresh data from it.
45
Some data gets requested repeatedly, and that is where caching comes in.
46
Imagine a request needs a piece of data.
47
First, the application checks the cache.
48
If the data is there, it returns it and that is cache hit.
49
If it is missing, the application fetches it from database
50
and can put it in the cache and that is a cache miss.
51
The next request may get the data without querying the database again.
52
But caching introduces a problem.
53
What happens when the original data changes?
54
The cached copy might become stale.
55
You can use techniques such as time to live or TTL to expire entries.
56
You can also use invalid cache data when the underlying data changes.
57
Strategies like write -through or write -around handle writes differently.
58
And there is no single strategy that's best for every workload.
59
Uber has built caching infrastructure called CacheFront.
60
In August 1995, Uber reported that it handled more than 150 million cache row reads per second at peak.
61
Notice what that number means.
62
It's cache database row reads handled by that infrastructure.
63
And it does not mean every database read across
64
Uber was subbed from the cache.
65
Caching can remove a lot of repeated work, but it does not solve every database problem.
66
Some applications read data far more often than updated.
67
For those workloads, you can introduce read replicas.
68
The primary database handle writes, replica receive copies of the data then application can send suitable read queries to those replicas
69
that take some read traffic away from the primary
70
but there is an important catch replication can take time the primary might have the latest update
71
while the replica is still catching up and this is called replication lack
72
so if a request need the fresh data you may need a read from the primary.
73
Replica help with re -capacity but they don't automatically solve right bottlenecks or storage limits.
74
Now imagine your trip data keeps growing.
75
Eventually one database may struggle to store and serve everything efficiently.
76
Uber ran into storage challenges with the Postgres -packed trip data in its early growth period.
77
It developed a system called SchemaLess on top of MySQL.
78
This brings us to another scaling technique.
79
It is called Sharding.
80
Instead of storing all the data on one database, you split it across multiple database groups.
81
For example, imagine dividing trips into groups based on hash of the trip identifier.
82
One group stores some trips, another group stores others.
83
The application needs a way to work out which group contains the requested trip.
84
And that is the basic idea.
85
This example explains sharding.
86
It is not a claim about Uber's exact partitioning algorithm okay
87
so sharding lets you distribute data and workload
88
but now your application need to know where each record live
89
and the workload might not be evenly distributed imagine one shard receiving far more requests than the others
90
that shard can become a bottleneck even when the rest are healthy
91
and this is also called a hot shard so sluting the data helps but it also makes data management more complicated.
92
Okay, so there is one more challenge.
93
As the developer team grow, more developers work on the same application.
94
A change in one area might affect other.
95
Deployment can become much harder to coordinate.
96
One response is to split the application into small services.
97
And Uber documented its move toward a microservice architecture in 2015.
98
So imagine this.
99
Payments, pricing, and trips matching as separate services.
100
These are illustrative examples of how an application can be divided.
101
Each service can have its own code base and deployment process.
102
Teams can work more independently, but now one user request might depend on several services working together.
103
And that creates a different kind of problem.
104
Okay, inside one application, calling a function is usually straightforward.
105
For example, a fare calculation function might take the trip details as input and return the calculated fare.
106
But when one service calls another service over a network, things change.
107
The other service might be slow or it might be unavailable or the network request might time out.
108
Now you need to handle failure across service boundaries.
109
Okay, so when an application has many services, finding the cause of slow request gets hard.
110
Was the delay in pricing, was it payments or was one service waiting on another?
111
So Uber developed Yaeger, a distributed tracing system to help engineer follow requests across services.
112
Tracing helps you to see where time is spent and where request is failing.
113
Without that visibility, debugging a distributed system becomes much harder.
114
Ok, as the services multiply, service discovery and communication also becomes more complicated.
115
A service needs to find the service it wants to call.
116
And the system need to handle those connections reliably.
117
So Uber developed Hyperband as a part of its service communication infrastructure.
118
So engineers also used general resilience patterns such as timeouts, retries and circuit breakers.
119
But those patterns need careful configuration.
120
Retries for example can make an overload service even busier.
121
So by July 2020, Uber had described an architecture with around 2200 critical microservices organized into 70 domains.
122
These 70 domains were not 70 services.
123
They were broader grouping used to organize services.
124
And that is important lesson here.
125
Microservices can be more independent but also create more dependence, more network calls, more operational complexity.
126
So just look at the progression.
127
You start with one application, then you separate the database, you add application servers, you share session state, you introduce caching and read applicas,
128
you shard the data, and eventually you split the application into services.
129
Each decision solves a problem, but each decision also introduces a new trade -off.
130
That's how production systems grow.
131
Not by adding every technology from the beginning, but by identifying the bottleneck, choosing a solution, and understanding the new problem that solution create.
132
When you are designing a system, don't just ask, what technology should I use?
133
Ask, what problem am I solving and what trade -off am I accepting?
134
Save this for your next systemline interview and follow for more.

下載應用

流利口語所需的一切

AI 發音評分逐句打分
音標練習掌握每個音
詞彙學習建立你的詞庫
詞彙遊戲邊玩邊學
FOR YOU

推薦影片

本課的詞彙與口語要點

這節口語課以影片“How Uber Scaled from One Server to 150 Million Reads Per Second”為素材。 影片中反覆出現的詞有:application, database, server, request, Uber。 這段影片共有 134 個句子、1507 個單詞可供跟讀。 講話部分時長為 9:46。 說話人語速自然,每分鐘約 154 個詞,接近日常對話。 只有 79% 的單詞屬於英語最常用的 3,000 詞,詞彙難度較高。

影片中的重點詞彙

影片中 15 個值得學習的單詞,附發音和釋義:

單詞發音釋義
database 名詞/ˈdeɪtəˌbeɪs/資料庫 /資料庫
server 名詞/ˈsɝvɚ/伺服器 /伺服器, 伺服器
cache 名詞/kæʃ/快取 /快取, 快取
distribute 動詞/dɪˈstɹɪbjuːt/分配
replica 名詞/ˈɹɛplɪkə/複製品 /複製品, 仿製品 /仿製品
infrastructure 名詞/ˈɪnfɹəˌstɹʌkt͡ʃə/基礎設施 /基礎設施
solve 動詞/sɒlv/解決 /解決
introduce 動詞/ˌɪntɹəˈdus/介紹 /介紹, 紹介 /紹介
query 名詞/ˈkwɪɹ.i/問題 /問題, 質問 /質問
necessarily 副詞/ˌnɛs.əˈsɛɹ.ə.li/必然地, 必定地
complicated 形容詞/ˈkɑm.plɪˌkeɪ.tɪd/複雜 /複雜
technique 名詞/tɛkˈniːk/技術 /技術, 技巧
architecture 名詞/ˈɑː.kɪˌtɛk.t͡ʃə/建築 /建築, 建築學 /建築學
log 名詞/lɔɡ/圓木 /圓木
domain 名詞/ˌdəʊˈmeɪ̯n/領域 /領域

影片中出現的短語動詞

單詞發音釋義
look up 動詞/ˌlʊk ˈʌp/仰, 上看
work out 動詞/ˌwɝk ˈaʊt/計算 /計算, 算出

影片中的語法

說話人最常用的結構,並附上影片中的原話:

結構影片中的用法
被動語態 be + 過去分詞 — 強調發生了什麼,而不是誰做的is called · are logged · was subbed
現在完成時 have/has + 過去分詞 — 過去發生但與現在仍有關聯的事We've added · has built

需要注意的發音

說話人用了 6 次縮略和弱讀形式,例如 don't, we've。請按聽到的簡短形式來說。

  • “th” 音: algorithm /ˈælɡəɹɪðm̩/, unhealthy /ʌnˈhɛl.θi/, monolith /ˈmɑn.əˌlɪθ/
  • “sh” 和 “zh” 音: cache /kæʃ/, operational /ˌɑ.pəˈɹeɪ.ʃə.nəl/, configuration /kənˌfɪɡ.əˈɹeɪ̯.ʃən/, progression /pɹəˈɡɹɛʃən/, calculation /ˌkælkjuˈleɪʃn̩/
  • 長單詞——注意重音位置: infrastructure /ˈɪnfɹəˌstɹʌkt͡ʃə/, necessarily /ˌnɛs.əˈsɛɹ.ə.li/, complicated /ˈkɑm.plɪˌkeɪ.tɪd/, architecture /ˈɑː.kɪˌtɛk.t͡ʃə/, developer /dɪˈvɛləpɚ/

中文母語者容易讀錯的音:

  • /v/ — 不要讀成 /w/,上齒輕觸下唇: server /ˈsɝvɚ/, solve /sɒlv/, developer /dɪˈvɛləpɚ/, divide /dɪˈvaɪd/, unavailable /ˌʌnəˈveɪləbəl/
  • 詞尾子音 — 要讀清楚,後面不要加母音: database /ˈdeɪtəˌbeɪs/, cache /kæʃ/, distribute /dɪˈstɹɪbjuːt/, solve /sɒlv/, introduce /ˌɪntɹəˈdus/
  • /r/ 和 /l/ 的區別: replica /ˈɹɛplɪkə/, necessarily /ˌnɛs.əˈsɛɹ.ə.li/, differently /ˈdɪf.ə.ɹənt.li/, operational /ˌɑ.pəˈɹeɪ.ʃə.nəl/, repeatedly /ɹɪˈpiːtɪdli/

如何用這段影片練習

  1. 先完整聽一遍影片,不要開口,記下不認識的單詞。
  2. 先用 0.75 倍速逐句跟讀,熟練之後再回到正常速度。
  3. 錄下自己的聲音並與原聲對比,特別注意 database, server, cache 這類單詞。

什麼是跟讀法?

跟讀法 (Shadowing) 是一種有科學依據的語言學習技巧,最初開發用於專業口譯員的培訓,並由多語言者Alexander Arguelles博士普及。這個方法簡單而強大:您在聽英語母語原聲的同時立即大聲重複——就像是一個延遲1-2秒緊跟說話者的影子。與被動聽力或語法練習不同,跟讀法強迫您的大腦和口腔肌肉同時處理並模模擬實的講話模式。研究表明它能顯著提高發音準確性,語調,節奏,連讀,聽力理解和口語流利度——使其成為雅思口語備考和真實英語交流最有效的方法之一。

影子跟讀法: 閱讀完整分步指南 →