my-redis

Problem statememt-:

mini-redis

If you are not familiar with how Redis works – read up on https://redis.io

Create a service which do the following: -

•    Store a key value pair in memory via a http POST method
  

o For example, posting {‘k’: “v”} to http:///my-redis/ should store the value in the server and return the appropriate HTTP response code

•    Retrieve a value from memory given the key via a http GET method
  

o For example, a GET call on http:///my-redis/k should return “v” and return the appropriate HTTP response code

•    Retrieve an incremented value of a key via a http PUT method
  

o For example, posting {‘ik’: 1} to http:///my-redis/ should store the value in the server and return the appropriate HTTP response code o A PUT call to http:///my-redis/increment/ik should internally increment the value of the key ik in the server to 2 and return 2 along with the applicable response codes o A GET call to http:///my-redis/ik should return 2 along with the applicable http response code to prove the above point

Points to note:

•    Extra credit if you can do this in java or golang
  

•   Extra credit if you can do this  The increment method MUST be written in a thread-safe manner. 
  

•   To prove this, 
  

o Post {‘threadSafeKey’: 1} to your server http:///my-redis o Run apache bench -- ab -n 100 -c 10 http:///my-redis/increment/threadSafekey (100 calls with a max of 10 being concurrent) o A GET call on http:///my-redis/threadSafeKey should return 101 along with the appropriate http response code

Solution proposed-:

There are three flavours of Update operations-:

1. MyRedisMapComputeDao - This uses ConcurrentHashMap's computeIfPresent to atomically update.
2. MyRedisAtomicCASUpdateDao - This uses AtomicInteger for CHM value, in high contention, it may waste CPU cycles
3. MyRedisConcurrentUpdatePerformantDao - This uses LongAdder which is recommended for high concurrent updates(high contention).

Apache Bench was used to measure the performance of these three flavors in concurrency level of 10 with 1000 requests(AB is not for load test, please use JMeter) using following-:

ab -u empty_file.txt -n 1000 -c 10 http://127.0.0.1:8080/my-redis/key/

And below are the results -:

---

For myRedisMapComputeDao

Finished 1000 requests

Server Software:
Server Hostname: 127.0.0.1 Server Port: 8080

Document Path: /my-redis/key/ Document Length: 61 bytes

Concurrency Level: 10 Time taken for tests: 0.357 seconds Complete requests: 1000 Failed requests: 0 Total transferred: 166000 bytes Total body sent: 140000 HTML transferred: 61000 bytes Requests per second: 2802.67 [#/sec] (mean) Time per request: 3.568 [ms] (mean) Time per request: 0.357 [ms] (mean, across all concurrent requests) Transfer rate: 454.34 [Kbytes/sec] received 383.18 kb/s sent 837.52 kb/s total

Connection Times (ms) min mean[+/-sd] median max Connect: 0 1 0.7 0 6 Processing: 0 3 6.1 2 56 Waiting: 0 2 6.1 1 56 Total: 1 3 6.1 2 57 WARNING: The median and mean for the initial connection time are not within a normal deviation These results are probably not that reliable.

Percentage of the requests served within a certain time (ms) 50% 2 66% 3 75% 3 80% 3 90% 4 95% 6 98% 29 99% 56 100% 57 (longest request)

For myRedisAtomicCASUpdateDao

Finished 1000 requests

Server Software:
Server Hostname: 127.0.0.1 Server Port: 8080

Document Path: /my-redis/key/ Document Length: 61 bytes

Concurrency Level: 10 Time taken for tests: 0.329 seconds Complete requests: 1000 Failed requests: 0 Total transferred: 166000 bytes Total body sent: 140000 HTML transferred: 61000 bytes Requests per second: 3041.00 [#/sec] (mean) Time per request: 3.288 [ms] (mean) Time per request: 0.329 [ms] (mean, across all concurrent requests) Transfer rate: 492.97 [Kbytes/sec] received 415.76 kb/s sent 908.74 kb/s total

Connection Times (ms) min mean[+/-sd] median max Connect: 0 1 0.6 1 5 Processing: 0 2 3.6 2 32 Waiting: 0 2 3.6 1 32 Total: 1 3 3.6 2 33

Percentage of the requests served within a certain time (ms) 50% 2 66% 3 75% 3 80% 4 90% 5 95% 6 98% 18 99% 32 100% 33 (longest request)

---

For myRedisConcurrentUpdatePerformantDao

Finished 1000 requests

Server Software:
Server Hostname: 127.0.0.1 Server Port: 8080

Document Path: /my-redis/key/ Document Length: 61 bytes

Concurrency Level: 10 Time taken for tests: 0.342 seconds Complete requests: 1000 Failed requests: 0 Total transferred: 166000 bytes Total body sent: 140000 HTML transferred: 61000 bytes Requests per second: 2919.84 [#/sec] (mean) Time per request: 3.425 [ms] (mean) Time per request: 0.342 [ms] (mean, across all concurrent requests) Transfer rate: 473.33 [Kbytes/sec] received 399.20 kb/s sent 872.53 kb/s total

Connection Times (ms) min mean[+/-sd] median max Connect: 0 1 0.6 1 5 Processing: 1 2 1.4 2 12 Waiting: 0 2 1.3 2 11 Total: 1 3 1.5 3 13

Percentage of the requests served within a certain time (ms) 50% 3 66% 3 75% 4 80% 4 90% 5 95% 7 98% 8 99% 9 100% 13 (longest request)

---

Conclusion -: In concurrent environment, all variants(myRedisMapComputeDao, myRedisAtomicCASUpdateDao, myRedisConcurrentUpdatePerformantDao) performed in a thread safe manner but there was almost no performance difference seen across the three variants for update, which needs more figuring