You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on May 5, 2023. It is now read-only.
When running a simple script that responds an HTTP request with tableService.insertOrReplaceEntity, on a single core machine (Azure), my maximum throughput is just 48 req/sec, with the CPU at 10% (in other words, CPU is very idle).
When running the same script under node-cluster, in the same single core machine but now with 4 node processes, throughput reaches 134 req/sec.
Issue 1 - Is Azure SDK for node really 100% asynchronous? This data seems to indicate it is not, but I can't verify. node-cluster should not make a difference, since this is a single core machine.
Issue 2 - Even with node-cluster, throughput is too low, since the Azure documentation indicates a limit of 2000 entities per second in a single partition. Why?
When running a simple script that responds an HTTP request with tableService.insertOrReplaceEntity, on a single core machine (Azure), my maximum throughput is just 48 req/sec, with the CPU at 10% (in other words, CPU is very idle).
When running the same script under node-cluster, in the same single core machine but now with 4 node processes, throughput reaches 134 req/sec.
Issue 1 - Is Azure SDK for node really 100% asynchronous? This data seems to indicate it is not, but I can't verify. node-cluster should not make a difference, since this is a single core machine.
Issue 2 - Even with node-cluster, throughput is too low, since the Azure documentation indicates a limit of 2000 entities per second in a single partition. Why?
#1230
The text was updated successfully, but these errors were encountered: