Github user michaelandrepearce commented on a diff in the pull request:
https://github.com/apache/activemq-artemis/pull/2401#discussion_r229089174
--- Diff:
artemis-server/src/main/java/org/apache/activemq/artemis/core/paging/impl/PagingManagerImpl.java
---
@@ -437,12 +449,27 @@ public void processReload() throws Exception {
}
}
+ private boolean ignoreGlobalMaxSize(SimpleString address) {
+ if (this.addressPrefixesIgnoringGlobalMaxSize == null) {
+ return false;
+ } else {
+ //for a small number of prefixes to check we can use just a
linear search too :)
+ for (SimpleString prefixes :
this.addressPrefixesIgnoringGlobalMaxSize) {
--- End diff --
why even code this to be more than one, concern here would be what else?
could end up in non-managed space? Its a dangerous thing to have anything
unbounded entirely. See my comment about possible alternative and safer
approach with implementing a two tier global max sizing, one global (all
addresses) and one non-management max which would be set lower to allow head
room for the management ones no matter what (all addresses except management)
---
