Mark Payne created NIFI-13796:
---------------------------------
Summary: Improve efficiency of FlowFile content slab allocation
Key: NIFI-13796
URL: https://issues.apache.org/jira/browse/NIFI-13796
Project: Apache NiFi
Issue Type: Improvement
Reporter: Mark Payne
Assignee: Mark Payne
Fix For: 2.0.0-M5
When we write to the content repository, we write until the we hit the
configurable max appendable claim size. Then we move on to another resource
claim in the content repository. However, this only happens inter-session, as
the session holds open an OutputStream to a Content Claim for efficiency
purposes.
This, however, has the downside of being less efficient in terms of splitting
FlowFile across content claims. As a result, we can write significantly far
beyond the 50 KB max when writing many FlowFiles in a session (think
UnpackContent, Split*, etc.)
We can improve this by checking the size of the content claim + offset when we
finish writing to a content claim and only adding back to our active claim
queue if that size is less than the max appendable size. This should improve
NiFi's ability to cleanup content more aggressively.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
