Only poll work when there are slots to work on

[cschleiden]

This pull request refactors the worker queue implementation to introduce a new workQueue abstraction, improving the management of task concurrency and queue operations. The main changes include replacing the direct use of channels for task queuing and concurrency control with the new workQueue type, updating the worker logic to use this abstraction, and adding comprehensive tests for the new queue.

If the number of max parallel tasks a worker can work is constrained, it will pause polling until capacity is available.

[cschleiden]

!bench

[github-actions]

Benchmark Results

MySQL run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
mysql-main	533.2 ± 202.9	404.1	1094.6	1.06 ± 0.43
mysql-pr	502.4 ± 66.4	391.5	586.0	1.00

SQLite run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
sqlite-main	510.1 ± 89.5	393.3	651.5	1.00
sqlite-pr	520.5 ± 103.9	375.3	643.2	1.02 ± 0.27

Redis run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
redis-main	190.4 ± 21.7	160.5	246.4	1.01 ± 0.19
redis-pr	189.2 ± 28.1	130.4	251.7	1.00

[cschleiden]

!bench large

[github-actions]

Benchmark Results

MySQL run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
mysql-main	500.5 ± 52.8	444.7	627.3	1.00
mysql-pr	568.1 ± 50.1	489.3	642.5	1.13 ± 0.16

SQLite run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
sqlite-main	714.7 ± 131.8	460.9	843.8	1.00
sqlite-pr	737.8 ± 107.6	614.3	864.8	1.03 ± 0.24

Redis run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
redis-main	195.6 ± 31.9	152.6	256.3	1.04 ± 0.21
redis-pr	188.2 ± 22.4	172.3	250.3	1.00

[cschleiden]

!bench verylarge

[github-actions]

Benchmark Results

MySQL run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
mysql-main	560.7 ± 86.6	441.5	701.3	1.03 ± 0.23
mysql-pr	546.2 ± 91.0	429.4	681.5	1.00

SQLite run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
sqlite-main	633.8 ± 120.3	434.6	838.9	1.07 ± 0.32
sqlite-pr	592.3 ± 139.0	414.8	822.9	1.00

Redis run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
redis-main	198.5 ± 18.1	170.7	247.2	1.10 ± 0.20
redis-pr	180.7 ± 27.6	130.5	249.3	1.00

[cschleiden]

!bench !large

[github-actions]

Benchmark Results

MySQL run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
mysql-main	601.6 ± 107.2	467.8	851.5	1.11 ± 0.31
mysql-pr	541.2 ± 114.2	457.1	860.3	1.00

Large MySQL payload run (1MB)

Command	Mean [s]	Min [s]	Max [s]	Relative
mysql-main	24.968 ± 1.006	23.284	26.494	1.00
mysql-pr	30.961 ± 18.858	24.131	84.610	1.24 ± 0.76

SQLite run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
sqlite-main	626.0 ± 128.4	425.4	849.1	1.02 ± 0.27
sqlite-pr	614.9 ± 104.2	415.5	810.6	1.00

Large SQLite payload run (1MB)

Command	Mean [s]	Min [s]	Max [s]	Relative
sqlite-main	22.388 ± 0.695	20.997	23.107	1.00
sqlite-pr	22.664 ± 0.738	21.316	23.480	1.01 ± 0.05

Redis run

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
redis-main	208.3 ± 25.4	185.9	265.8	1.00
redis-pr	211.6 ± 31.1	170.9	254.5	1.02 ± 0.19

Large Redis payload run (1MB)

Command	Mean [s]	Min [s]	Max [s]	Relative
redis-main	22.232 ± 0.327	21.803	22.746	1.01 ± 0.02
redis-pr	22.024 ± 0.406	21.507	22.587	1.00