Re: [I] [Bug] RoundRobinLoadBalance sends disproportionate traffic to newly scaled-up pods during rapid scaling in Dubbo 3.3 [dubbo]

Tue, 18 Nov 2025 21:59:14 -0800 


RainYuY commented on issue #15762:
URL: https://github.com/apache/dubbo/issues/15762#issuecomment-3550935776

   > I suspect the issue lies here: every time the weight changes, current is 
reset to 0. During the 5-minute warm-up period, the provider's weight keeps 
changing, causing each calculation to be set to 0.
   > 
   > <img alt="Image" width="1190" height="286" 
src="https://private-user-images.githubusercontent.com/20431373/516061275-2a1f1129-464d-48be-8f52-d49aaa799397.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.tzSIDBIuqiqzF9BG4XjDnSE0e6jVKcvcfHtwEzBMx0c";>
 <img alt="Image" width="992" height="312" 
src="https://private-user-images.githubusercontent.com/20431373/516061326-69332d4a-b92b-4959-8353-586339546012.png?jwt=eyJ
 
0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.wdiWK4QKiygLX-dTnMO-bfF-dYMYPKFPZA7uZIyD-94">
   > Here is our reasoning process.
   > 
   > <img alt="Image" width="957" height="581" 
src="https://private-user-images.githubusercontent.com/20431373/516061399-8d7cae09-8500-4a5f-8d64-ab39f2bcc96b.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.UZ4skalMRpXZfFeoq3l4mgnh2wyL12jIG925R1yoZ68";>
 <img alt="Image" width="1119" height="1069" 
src="https://private-user-images.githubusercontent.com/20431373/516061458-084967bb-1dde-4f1a-9f33-f9ce22ccac68.png?jwt=ey
 
J0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.aOHeU0nEd5b8kR4k0k0hZWZb6VdrdYXvR6vhR80-TPg">
 <img alt="Image" width="1259" height="819" 
src="https://private-user-images.githubusercontent.com/20431373/516061513-e15b74b0-8b4c-4984-b7ec-d2780e6500d0.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NjM1MzIwMjAsIm5iZiI6MTc2MzUzMTcyM
 
CwicGF0aCI6Ii8yMDQzMTM3My81MTYwNjE1MTMtZTE1Yjc0YjAtOGI0Yy00OTg0LWI3ZWMtZDI3ODBlNjUwMGQwLnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNTExMTklMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjUxMTE5VDA1NTUyMFomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPTY4N2FjOWMxMWIyMTM3MGYyOTgxOGQ2OTNlNjEyYjIzNGUyMGFhNWI2OTExYjM4OTk0NjQ2MDE2YmFlZmVhY2QmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0In0.aaFRZxrOsIG6Z89ooeGX4vpI86WPf6p3Nnkftfg5Q2A">
 <img alt="Image" width="1100" height="1049" 
src="https://private-user-images.githubusercontent.com/20431373/516061574-8f637c5b-743d-46f4-9891-720b6d2de7ee.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NjM1MzIwMjAsIm5iZiI6MTc2MzUzMTcyMCwicGF0aCI6Ii8yMDQzMTM3My81MTYwNjE1NzQtOGY2MzdjNWItNzQzZC00NmY0LTk4OTEtNzIwYjZkMmRlN2VlLnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFW
 
Q09EWUxTQTUzUFFLNFpBJTJGMjAyNTExMTklMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjUxMTE5VDA1NTUyMFomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPTRmOTY0ODdjZjQwZTBmYmFhNjRlZWU0YTQxMzkwMmUyNmNmNDYxMDE4ZTMyMzkxZTI4YWY0NDJjMGU2NDAwYTcmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0In0.OY_IdieQg3DWYmvAHK7511Ap9WvMFF0cfuI6vDUS6Pg">
   
   There's a slight difference. I don't think the old node will show a large 
negative number. Because the total weight during the warm-up phase won't be 
very large, and the total weight will also increase as the warm-up progresses.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to