Murtadha Hubail created ASTERIXDB-2276:

             Summary: Introduce max active writable datasets
                 Key: ASTERIXDB-2276
             Project: Apache AsterixDB
          Issue Type: Improvement
            Reporter: Murtadha Hubail
            Assignee: Murtadha Hubail

currently, it is possible for an operation that tries to allocates memory for 
its dataset to fail without the user knowing about it. For example, if a feed 
is connected to a number of datasets that exceeds the max global memory for 
datasets, this issue will happen. The proposal is to add a configurable max 
active writable datasets parameter and calculate the memory component size 
based on it and the the memory component page size. After that, any operation 
could check the currently modified datasets and fail early if that number 
exceeds the configured value.

This message was sent by Atlassian JIRA

Reply via email to