Hello Neelaksh Sheel,

I am sharing my perspective in a generic view.

1. Average Latency of the Service

2. Latency w.r.t No . of Users / Concurrent Users

3. Memory, CPU, SWAP, Disk, Virtual Memory Utilization w.r.t particular web service ( its basically a process running in a remote server). This enable to select a better infrastructure configuration.

4. Average Data Transfer during testing. (This helps to select a proper bandwidth plan. Some Application may require more incoming bandwidth, less  outgoing bandwidth and vice versa)

5. No. of successful and failure transactions.

6. CPU Temperature & Disk Temperature during testing (if your servers are SMART enabled)

Hope these metrics help you.

-

V. Srikrishnan

Hi Neelaksh Sheel,

  You can have a look at a systematic mapping of quality models for web services, where you will find the list of performance metrics (among others) and how they are used/defined.

https://www.researchgate.net/publication/261599651_Quality_models_for_web_services_A_systematic_mapping

You can also have a look at the additional resources provided at:

http://gessi.lsi.upc.edu/qmodels/ 

Please find below some of the metrics:

Time behaviour:

- Response time, 

- Execution time

- Network Latency

- queue delay time

- Jitter

- Throughput

Resource utilization:

- Memory used

- CPU used

- Bandwith

- Queue length

- Input size

- Output size

Capacity:

- Maximum throughput

- Load balancing

- Processor speed

- Memory capacity

- Maximum concurrent connection

Hope it helps :)

Best regards,

Marc.

Article Quality models for web services: A systematic mapping

Add Soak testing to the above list.