Executive Summary

Objective

To validate the abc.in performance under the load of 50000 transactions per hour and to identify the breaking point of the server.

Load Test Result

Load test has been conducted with 50 users concurrently accessing the website to achieve the 50000 transactions per hour. This test is to validate whether the server can handle 50000 transactions per hour or not.

Total Duration	1 hour

Label	# Samples	Average	Median	90% Line	95% Line	99% Line	Min	Max	Error %	Throughput	KB/sec
01_HomePage	9868	672	573	1177	1297	1904	290	6243	0.00%	2.6	27.2
02_Login	9866	404	290	927	959	1326	284	14944	0.00%	2.6	10.6
03_Contacts	9866	400	291	925	955	1300	283	5737	0.00%	2.6	10.6
04_Team	9863	409	292	931	963	1298	284	12049	0.00%	2.6	10.6
05_UserProfile	9863	403	291	926	964	1302	284	23931	0.00%	2.6	10.6
TOTAL	49326	458	292	935	993	1543	283	23931	0.00%	13	69.6

Observations:

1. No errors encountered during load testing
2. Maximum response time is 23931 milliseconds
3. Minimum response time is 283 milliseconds
4. 99 percentile is less 1543 milliseconds

Stress Testing

In this test, more than 100 users have been injected concurrently at a regular interval of 30 seconds for the below transactions. When the concurrent users reached 125, the first error (Internal Server Error Code 500) occurred. Eventually all the transactions started failing. Below is the response time dashboard for the run 1. Similar observations for the run 2 as well.

Run #1

Label	# Samples	Average	Median	90% Line	95% Line	99% Line	Min	Max	Error %	Throughput	KB/sec
01_HomePage	4068	1992	1365	3067	4240	11402	66	235596	0.64%	9.1	95.1
02_Login	4042	1417	948	2023	3428	12325	1	140924	0.59%	9.1	37
03_Contacts	4017	1459	975	1971	3299	11753	2	105727	0.67%	9.1	36.8
04_Team	3987	1289	963	1943	3289	10987	251	97139	0.53%	9	36.6
05_UserProfile	3965	1333	976	2041	3297	10619	165	121944	0.68%	9	36.4
TOTAL	20079	1500	1009	2343	3528	11507	1	235596	0.62%	45.1	241.3

Run #2

Label	# Samples	Average	Median	90% Line	95% Line	99% Line	Min	Max	Error %	Throughput	KB/sec
01_HomePage	5820	2435	1576	3590	5386	17901	1	365617	0.65%	9.2	96.3
02_Login	5780	1818	1007	2235	3777	16757	106	502888	0.64%	9.2	37.4
03_Contacts	5741	2006	1001	2233	3819	21007	1	303103	0.63%	9.1	37.2
04_Team	5705	1678	1006	2239	3833	17675	1	124876	0.49%	9.1	37
05_UserProfile	5675	1814	1001	2209	3642	18025	1	379085	0.62%	9.1	36.8
TOTAL	28721	1952	1033	2757	4324	18025	1	502888	0.61%	45.6	244.2

Observations:

1. No errors observed till 125 concurrent users performing the transactions.
2. First internal server error occurred after the 125 concurrent users trying to access the website.
3. System will eventually fail to respond after 125 concurrent users.

Conclusion

Multiple rounds have been conducted to validate the server performance and to identify the breaking point. During the load test, the system is capable of handling 50000 transactions per hour.

No errors have been observed during the execution which resembles that the server will respond positively for the load of 50000 transactions per hour with the 99 percentile response time is 1.5 seconds.

During the stress test, the server responded positively till 125 concurrent users, and when the load increases, the first internal server 500 error code occurred and eventually all the transactions started failing. This behavior has been observed in the subsequent execution as well.

Load test reveals that the server can handle 50000 transactions per hour with minimal users accessed the website.

Stress test reveals that the server can handle 125 concurrent users accessing the website. Subsequent increase in the load will fail to fulfill the user's expectation.

Next Steps

It is advisable to monitor the performance of CPU, Memory, Network and Disk utilization. To handle more number of users, tuning or scaling the infrastructure is required.