# Networking Metrics

## About

Networking metrics are essential performance indicators that help assess the efficiency, reliability, and responsiveness of a system’s network. They play a crucial role in designing scalable, high-performance distributed systems and optimizing network infrastructure.

## **1. Latency**

Latency is the **time delay** between when a request is made and when a response is received. It is measured in **milliseconds (ms)** and can impact system performance, user experience, and data transmission.

Latency can be classified into different types based on where it occurs:

1. Network Latency (End-to-End Delay)
2. Application Latency
3. Disk Latency (Storage Latency)
4. Database Latency
5. Memory Latency

### **1. Network Latency**

The delay in transmitting data over a network from sender to receiver.

#### **Causes of Network Latency**

* **Physical Distance** → Longer distances take more time (e.g., sending data from the US to India).
* **Congestion** → High traffic on a network slows down data transmission.
* **Packet Loss & Retransmission** → Lost packets require resending, increasing latency.
* **Bandwidth Limitations** → Lower bandwidth means slower data transfer.
* **Routing & Switching Delays** → Time spent at intermediate routers and switches.

#### **Formula for Total Network Latency**

`Network Latency = Transmission Delay + Propagation Delay + Processing Delay + Queuing Delay`

#### **Example Calculation**

* Packet Size = 10 MB
* Bandwidth = 100 Mbps
* Distance = 3000 km
* Speed of Light in Fiber = 200,000 km/s

Transmission Delay = (10×8) / 100 = 0.8 seconds\
Propagation Delay = 3000 / 200000 = 0.015 seconds\
\&#xNAN;*Total Latency ≈ 0.815 seconds (815 ms)*

#### **Optimization Techniques**

* Use **Content Delivery Networks (CDNs)** to bring data closer to users.
* Reduce **DNS resolution time** (avoid unnecessary redirects).
* Optimize **network routing** to use faster paths.

### **2. Application Latency**

The delay caused by an application while processing a request.

#### **Causes of Application Latency**

* **Slow Backend Processing** → Heavy computations or inefficient algorithms.
* **Database Queries** → Complex or unoptimized queries slow response time.
* **API Call Delays** → Waiting for responses from third-party APIs.
* **Thread Blocking** → Poor multi-threading can cause delays.

#### **Application Latency Breakdown**

<table data-header-hidden data-full-width="true"><thead><tr><th></th><th></th><th></th></tr></thead><tbody><tr><td><strong>Component</strong></td><td><strong>Definition</strong></td><td><strong>Optimization</strong></td></tr><tr><td><strong>Processing Time</strong></td><td>Time for backend logic execution</td><td>Use efficient algorithms</td></tr><tr><td><strong>Database Query Time</strong></td><td>Time spent fetching data from the DB</td><td>Optimize queries, indexing</td></tr><tr><td><strong>API Call Delay</strong></td><td>Delay from external APIs</td><td>Use async API calls, caching</td></tr></tbody></table>

#### **Formula for Application Latency**

`Application Latency = Processing Time + DB Query Time + API Call Delay`

#### **Example Calculation**

* Processing Time = 100 ms
* DB Query Time = 200 ms
* API Call Delay = 300 ms

*Total Latency = 100 + 200 + 300 = 600 ms*

#### **Optimization Techniques**

* **Reduce DB calls** → Use caching (Redis, Memcached).
* **Improve backend performance** → Optimize algorithms, use parallel processing.
* **Use Asynchronous Processing** → Reduce blocking operations.

### **3. Disk Latency (Storage Latency)**

The time delay in retrieving data from a storage device (HDD/SSD).

#### **Causes of Disk Latency**

* **Mechanical Movements in HDD** → Seek time, rotational delay.
* **Slow SSD Response** → NAND flash limitations.
* **Large File Transfers** → More data requires more time.

#### **Disk Latency Breakdown**

<table data-header-hidden data-full-width="true"><thead><tr><th width="245"></th><th width="407"></th><th></th></tr></thead><tbody><tr><td><strong>Component</strong></td><td><strong>Definition</strong></td><td><strong>Optimization</strong></td></tr><tr><td><strong>Seek Time</strong></td><td>Time for HDD head to reach the data location</td><td>Use SSD instead of HDD</td></tr><tr><td><strong>Rotational Delay</strong></td><td>Time waiting for the correct disk sector</td><td>Use NVMe SSDs</td></tr><tr><td><strong>Data Transfer Time</strong></td><td>Time to read/write data</td><td>Optimize storage architecture</td></tr></tbody></table>

#### **Formula for Disk Latency**

`Disk Latency=Seek Time + Rotational Delay + Data Transfer Time`

#### **Example Calculation**

* Seek Time = 5 ms
* Rotational Delay = 3 ms
* Data Transfer Time = 2 ms

*Total Latency = 5 + 3 + 2 = 10 ms*

#### **Optimization Techniques**

* Use **SSDs instead of HDDs**.
* Optimize **indexing and compression**.

### **4. Database Latency**

The delay in retrieving data from a database.

#### **Causes of Database Latency**

* **Slow Queries** → Poor indexing, large joins, unnecessary columns.
* **Locks and Contention** → Concurrent transactions blocking each other.
* **Network Bottlenecks** → Delays in transferring data from the database.

#### **Database Latency Breakdown**

<table data-header-hidden data-full-width="true"><thead><tr><th></th><th></th><th></th></tr></thead><tbody><tr><td><strong>Component</strong></td><td><strong>Definition</strong></td><td><strong>Optimization</strong></td></tr><tr><td><strong>Query Execution Time</strong></td><td>Time taken to run a query</td><td>Optimize queries</td></tr><tr><td><strong>Indexing Delay</strong></td><td>Time spent searching without indexes</td><td>Use proper indexing</td></tr><tr><td><strong>Lock Contention</strong></td><td>Delays due to concurrent transactions</td><td>Use optimistic locking</td></tr></tbody></table>

#### **Optimization Techniques**

* Use **indexes** to speed up lookups.
* Avoid **full-table scans**.
* Optimize **schema design** to reduce joins.

### **5. Memory Latency**

The delay in retrieving data from RAM.

#### **Causes of Memory Latency**

* **Cache Misses** → Data not found in L1/L2 cache.
* **Slow RAM Speeds** → Older DDR generations have higher latency.
* **Memory Bus Contention** → Multiple processes accessing memory simultaneously.

#### **Optimization Techniques**

* Use **faster RAM (DDR5 vs DDR4 vs DDR3)**.
* Optimize **memory allocation** (avoid excessive swapping).

## **2. Bandwidth**

The maximum rate of data transfer over a network connection in a given period.

### **Formula**

`Bandwidth = Total Data Transferred / Time Taken`

{% hint style="success" %}
**Throughput vs Bandwidth**: Bandwidth is the **theoretical limit**, whereas **throughput** is the actual amount of data transferred.

**Bandwidth Throttling**: Intentional reduction of bandwidth to control network congestion.
{% endhint %}

### **Optimization Techniques**

* Use **data compression** (e.g., Gzip, Brotli for web pages).
* Optimize **TCP window size** for large data transfers.
* Implement **QoS (Quality of Service) policies** to prioritize critical traffic.

## **3. Throughput**

The actual amount of data successfully transmitted per unit time.

### **Formula**

`Throughput = Total Successful Data Received / Time Taken`

### **Factors Affecting Throughput**

* **Network Congestion** – Delay caused by excessive traffic overwhelming network bandwidth.
* **Packet Loss** – Data packets failing to reach their destination due to network issues.
* **Hardware Limitations** – Performance bottlenecks due to outdated or underpowered network devices.

### **Optimization Techniques**

* Use **Load Balancing** to distribute traffic.
* Reduce **network bottlenecks** (e.g., upgrade network hardware).
* Implement **TCP optimization** (e.g., TCP Fast Open, congestion control algorithms).

## **4. Packet Loss**

The percentage of data packets that fail to reach their destination.

### **Formula**

`Packet Loss(%) = (Lost Packets / Total Packets Sent) × 100`

### **Causes**

* **Network Congestion** – Delay caused by excessive traffic overwhelming network bandwidth.
* **Faulty Hardware** – Malfunctioning network devices leading to degraded performance or connectivity issues.
* **Misconfigured Routers/Switches** – Incorrect network settings causing inefficient routing, bottlenecks, or connectivity failures.

### **Optimization Techniques**

* Use **redundancy mechanisms** like **Forward Error Correction (FEC)**.
* Implement **packet prioritization** for critical data.
* Optimize **MTU (Maximum Transmission Unit) size** to prevent fragmentation.

## **5. Jitter**

Variation in latency over time, causing delays in data packet arrival.

### **Formula**

`Jitter = ∣T2−T1∣`

where **T1** and **T2** are arrival times of two consecutive packets.

### **Impact**

* Causes **poor voice/video quality** in real-time applications (VoIP, Zoom, gaming).
* Leads to **buffering issues** in streaming services.

### **Optimization Techniques**

* Use **Jitter Buffers** to smooth packet delivery.
* Implement **Traffic Shaping** to prioritize real-time data.
* Reduce **network congestion** using QoS techniques.

## **6. Connection Establishment Time**

The time taken to establish a **network connection** (e.g., TCP handshake).

### **Formula**

`Connection Time = DNS Resolution Time + TCP Handshake Time + TLS Handshake Time`

### **Optimization Techniques**

* Use **persistent connections** (e.g., keep-alive in HTTP/2).
* Implement **TLS session resumption** to speed up HTTPS connections.
* Optimize **DNS resolution** using caching and fast resolvers.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://www.pranaypourkar.co.in/the-programmers-guide/system-design/performance-engineering/networking-metrics.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.