问题
I have an app service in Azure. It shows two metrics called Average Memory Working Set and Memory Working Set. Now, Memory Working Set is defined as the set of memory pages touched recently by the threads in the process. The graph for these two as shown in the portal are as below:
Now, there are three questions that I have:
- How do I find out what is the dedicated maximum memory my service has? What will happen once the limit is reached?
Memory Working Set is the number of memory pages touched recently by the threads in the process. I think that this means that the memory working set will only increase if my code has some memory leak, causes other pages to be loaded in memory etc. The question that I have is whether any external factors like number of requests have any effect on the Memory Working Set i.e. if the requests went up from 200 to 500, will the Memory Working Set increase? If yes, why?
How is Average Memory Working Set calculated? Is it calculated over time? Looking at the graph I see that the Average Memory Working Set and Memory Working Set have almost similar values.
回答1:
Let me try to answer as much as I can:
1.a To know dedicated maximum memory, you should check "private working set" for the process. I guess you can find this information in task manager or with help of KUDU(in Azure). Do some google, you'll find it easily.
1.b Once the limit is reached process will access Shared working set.
2.a Memory working set is 'Current amount of memory used by the app in MiBs'.
2.b If request went up by 2.5x times as you quoted, then Memory working set is most likely to increase as new object would get created, provided that there is shareable physical memory that can be allocated.
3.a Average Working Memory Set is 'The average amount of memory in MiBs used by the app'
来源：https://stackoverflow.com/questions/37299010/average-memory-working-set-and-memory-working-set-in-azure