Support server level consumption throttle

[xiangfu0]

Support server level consumption throttle:

1. Add new server config: to allow throttle realtime consumption speed at server level
2. Add new metric to count on server level realtime consumption rate.
3. Ratelimiter is based on the metric from 2.
4. Server Ratelimiter is one per RealtimeConsumptionRateManager, which is inside the InstanceHolder.

Release Notes:

Support server level consumption throttle:
- Add new server config: `pinot.server.consumption.rate.limit` to allow throttle realtime consumption speed at server level
- Default limit is 0, which means the server rate limiter is disabled.

[codecov-commenter]

Codecov Report

Attention: 4 lines in your changes are missing coverage. Please review.

Comparison is base (35faeb6) 61.58% compared to head (9be0311) 61.60%.
Report is 1 commits behind head on master.

Files	Patch %	Lines
...g/apache/pinot/common/metrics/AbstractMetrics.java	0.00%	3 Missing ⚠️
.../pinot/server/starter/helix/BaseServerStarter.java	0.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##             master   #12292      +/-   ##
============================================
+ Coverage     61.58%   61.60%   +0.01%     
+ Complexity     1153     1152       -1     
============================================
  Files          2417     2417              
  Lines        131611   131631      +20     
  Branches      20317    20318       +1     
============================================
+ Hits          81057    81087      +30     
+ Misses        44613    44599      -14     
- Partials       5941     5945       +4     

Flag	Coverage Δ
custom-integration1	<0.01% <0.00%> (ø)
integration	<0.01% <0.00%> (ø)
integration1	<0.01% <0.00%> (ø)
integration2	0.00% <0.00%> (ø)
java-11	61.54% <82.60%> (+0.01%)	⬆️
java-21	61.48% <82.60%> (+0.02%)	⬆️
skip-bytebuffers-false	61.57% <82.60%> (+<0.01%)	⬆️
skip-bytebuffers-true	61.45% <82.60%> (+0.02%)	⬆️
temurin	61.60% <82.60%> (+0.01%)	⬆️
unittests	61.59% <82.60%> (+0.01%)	⬆️
unittests1	46.71% <86.36%> (+0.02%)	⬆️
unittests2	27.73% <0.00%> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[snleee]

Thank you for adding this!

LGTM!

QQ: This will require the server restart to change the value. Is my understanding correct?

[sajjad-moradi]

LGTM. A couple of comments though:

1. Can you please update the main class documentation to reflect the new addition?
2. The additional logic is straight-forward. Do we really need an integration test? Having more tests in general is good, but this test adds a few minutes to the runtime of the test suits, and I'm not sure if it's worth incurring the cost.

[xiangfu0]

Thank you for adding this!

LGTM!

QQ: This will require the server restart to change the value. Is my understanding correct?

True, we can support the runtime change in the next PR.

[xiangfu0]

LGTM. A couple of comments though:

1. Can you please update the main class documentation to reflect the new addition?
2. The additional logic is straight-forward. Do we really need an integration test? Having more tests in general is good, but this test adds a few minutes to the runtime of the test suits, and I'm not sure if it's worth incurring the cost.

1. will do
2. I feel it's necessary for this one and we may want to add more tests later on for bytes-based throttling etc.
   Also we should take a look and deprecate/consolidate some tests if possible.