Reduce getNodeByQuery overhead

[fcostaoliveira]

The following PR does the following changes based upon on CPU profile info. The getNodeByQuery function represents 8.2% of an overhead of 12.3% when comparing single shard cluster with standalone.
Proposed changes:

• inlinging keyHashSlot to reduce overhead of that function call
• Reduce duplicate calls to getCommandFlags within getNodeByQuery
• moving crc16 to header file. inlining crc16 to reduce overhead of that function call

The above changes represent an improvement of approximately 5% on the achievable ops/sec as showcase bellow.

results

steps to reproduce

2 Nodes.
DB node (swap ip) :

taskset -c 0 ./src/redis-server --save '' --cluster-announce-ip 192.168.1.200 --bind 192.168.1.200 --protected-mode no --requirepass perf --daemonize yes --cluster-enabled yes --logfile redis.log
redis-cli -h 192.168.1.200 -a perf flushall
redis-cli -h 192.168.1.200 -a perf cluster flushslots
redis-cli -h 192.168.1.200 -a perf cluster addslotsrange 0 16383 
redis-cli -h 192.168.1.200 -a perf cluster info

Client node:

taskset -c 0,1 memtier_benchmark --server 192.168.1.200 --port 6379 --authenticate perf --cluster-mode --pipeline 10 --data-size 100 --ratio 1:0 --key-pattern P:P --key-minimum=1 --key-maximum 1000000 --test-time 180 -c 25 -t 2 --hide-histogram

results unstable (804110a)

root@micro1:~# taskset -c 0,1 memtier_benchmark --server 192.168.1.200 --port 6379 --authenticate perf --cluster-mode --pipeline 10 --data-size 100 --ratio 1:0 --key-pattern P:P --key-minimum=1 --key-maximum 1000000 --test-time 180 -c 25 -t 2 --hide-histogram
Writing results to stdout
[RUN #1] Preparing benchmark client...
[RUN #1] Launching threads now...
[RUN #1 100%, 180 secs]  0 threads:   107876070 ops,  599582 (avg:  599300) ops/sec, 83.99MB/sec (avg: 83.95MB/sec),  0.83 (avg:  0.83) msec latency

2         Threads
25        Connections per thread
180       Seconds


ALL STATS
======================================================================================================================================================
Type         Ops/sec     Hits/sec   Misses/sec    MOVED/sec      ASK/sec    Avg. Latency     p50 Latency     p99 Latency   p99.9 Latency       KB/sec 
------------------------------------------------------------------------------------------------------------------------------------------------------
Sets       599300.19          ---          ---         0.00         0.00         0.83133         0.82300         1.25500         3.07100     85967.12 
Gets            0.00         0.00         0.00         0.00         0.00             ---             ---             ---             ---         0.00 
Waits           0.00          ---          ---          ---          ---             ---             ---             ---             ---          --- 
Totals     599300.19         0.00         0.00         0.00         0.00         0.83133         0.82300         1.25500         3.07100     85967.12 

results this PR (7e1d2ea)

root@micro1:~# taskset -c 0,1 memtier_benchmark --server 192.168.1.200 --port 6379 --authenticate perf --cluster-mode --pipeline 10 --data-size 100 --ratio 1:0 --key-pattern P:P --key-minimum=1 --key-maximum 1000000 --test-time 180 -c 25 -t 2 --hide-histogram
Writing results to stdout
[RUN #1] Preparing benchmark client...
[RUN #1] Launching threads now...
[RUN #1 100%, 180 secs]  0 threads:   112251930 ops,  618650 (avg:  623611) ops/sec, 86.66MB/sec (avg: 87.36MB/sec),  0.80 (avg:  0.80) msec latency

2         Threads
25        Connections per thread
180       Seconds


ALL STATS
======================================================================================================================================================
Type         Ops/sec     Hits/sec   Misses/sec    MOVED/sec      ASK/sec    Avg. Latency     p50 Latency     p99 Latency   p99.9 Latency       KB/sec 
------------------------------------------------------------------------------------------------------------------------------------------------------
Sets       623611.19          ---          ---         0.00         0.00         0.79897         0.79100         1.20700         2.76700     89454.70 
Gets            0.00         0.00         0.00         0.00         0.00             ---             ---             ---             ---         0.00 
Waits           0.00          ---          ---          ---          ---             ---             ---             ---             ---          --- 
Totals     623611.19         0.00         0.00         0.00         0.00         0.79897         0.79100         1.20700         2.76700     89454.70 

[jhershberg]

While we're at it, do we want to put an "unlikely" here? I am far from an expert, just asking.

[fcostaoliveira]

will to the test of the impact on the results and reply back.