# Experiment 9 – SRE Golden Signals (LETS) Monitoring

Objective: Evaluate the reliability of a university examination portal using the four SRE Golden Signals: Latency, Traffic, Errors, and Saturation.

## Viva Questions and Answers

1\. What is SRE?

SRE stands for Site Reliability Engineering. It applies software engineering principles to IT operations to improve the reliability, availability, and performance of services.

2\. What are the four Golden Signals of SRE?

The four Golden Signals are:

- Latency: Time taken to respond to a request.
- Traffic: Number of requests received by the service.
- Errors: Number or percentage of failed requests.
- Saturation: How heavily the system resources are being utilized.

3\. What is latency?

Latency is the time taken by a system to process a request and return a response. High latency indicates slow service performance. Successful and failed requests should be analyzed separately because failed requests can sometimes return very quickly.

4\. What is traffic in monitoring?

Traffic represents the demand on a service, usually measured in requests per second (RPS). A sudden increase in traffic may overload the server and affect response time.

5\. What are errors in service monitoring?

Errors are requests that fail or produce incorrect responses. HTTP 5xx errors generally indicate server-side failures, while 4xx errors generally indicate client-side issues.

6\. What is saturation?

Saturation indicates how close a system is to reaching its resource limits. It can be measured using CPU utilization, memory usage, disk I/O, and request queue length.

7\. How do you measure response time using `curl`?

The `curl` command can measure different stages of an HTTP request using the `-w` option.

```
curl -o /dev/null -s -w \
"DNS: %{time_namelookup}\nTCP: %{time_connect}\nTLS: %{time_appconnect}\nTTFB: %{time_starttransfer}\nTotal: %{time_total}\n" \
https://example.com
```

8\. What is DNS lookup time?

It is the time taken to resolve a domain name into an IP address. High DNS lookup time can indicate DNS resolution delays.

9\. What are TCP connection time and TLS handshake time?

- TCP connection time: Time taken to establish a TCP connection with the server.
- TLS handshake time: Time taken to establish a secure HTTPS connection.

10\. What is TTFB?

TTFB stands for Time To First Byte. It measures the time from the start of the request until the first byte of the response is received. It helps identify server processing and network delays.

11\. How do you calculate request rate?

Request rate measures how many requests the server receives per second.

\\[ \text{Request Rate}=\frac{\text{Total Requests}}{\text{Time Interval}} \\]

For example, 600 requests in 60 seconds gives 10 requests per second.

12\. How do you calculate HTTP error rate?

Error rate is the percentage of failed requests out of total requests.

\\[ \text{Error Rate}= \frac{\text{Failed Requests}}{\text{Total Requests}}\times100 \\]

For example, if 50 out of 1000 requests fail, the error rate is 5%.

13\. What are SLI and SLO?

- SLI (Service Level Indicator): A measurable metric used to evaluate service reliability, such as request success rate or latency.
- SLO (Service Level Objective): The target value set for an SLI over a specific time period.

Example: An SLO can require 99.9% successful requests over 30 days.

14\. What is the difference between SLI, SLO, and SLA?

- SLI: What we measure.
- SLO: The reliability target we aim to achieve.
- SLA: A formal agreement with users or customers that may include consequences if targets are not met.

15\. How would you identify service degradation using the Golden Signals?

I would check all four signals:

- High latency indicates slow responses.
- High traffic indicates increased demand.
- High errors indicate failed requests.
- High saturation indicates resource exhaustion.

For example, high traffic combined with high CPU utilization and increasing response times may indicate server overload.

## Last-Minute Revision

| Term       | Meaning                              |
| ---------- | ------------------------------------ |
| SRE        | Site Reliability Engineering         |
| LETS       | Latency, Errors, Traffic, Saturation |
| Latency    | Time taken to respond                |
| Traffic    | Requests per second                  |
| Errors     | Failed requests                      |
| Saturation | Resource utilization                 |
| DNS        | Domain name resolution               |
| TCP        | Connection establishment             |
| TLS        | Secure connection handshake          |
| TTFB       | Time to First Byte                   |
| SLI        | Measurable reliability indicator     |
| SLO        | Target for an SLI                    |
| SLA        | Formal service agreement             |
| RPS        | Requests per second                  |

## Important for Practical Viva

Focus especially on:

1. The four Golden Signals and their meanings.
2. `curl` timing parameters: `time_namelookup`, `time_connect`, `time_appconnect`, `time_starttransfer`, `time_total`.
3. Request rate and error rate formulas.
4. Difference between SLI, SLO, and SLA.
5. Identifying service degradation using latency, errors, traffic, and saturation.
6. Recommending an alert for sustained high error rate or latency rather than a momentary spike.

Based on Experiment 9 of your CS414 lab question bank.&#x20;

CS414_ITIMM_Lab_ESE.pdf
