DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://issues.apache.org/bugzilla/show_bug.cgi?id=25372>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE.
http://issues.apache.org/bugzilla/show_bug.cgi?id=25372 auto close idle connections ------- Additional Comments From [EMAIL PROTECTED] 2004-04-19 16:23 ------- > 1) Looks like IdleConnectionHandler will have multi-threading issues: the add > method modifies the tree, while the closeIdleConnections method is trying to > iterate through the keySet. I think it perfectly OK for IdleConnectionHandler to assume that multi-threading issues will be taken care of by the HttpConnectionManager > 2) I'm not sure I understand why IdleConnectionHandler has a tree of > lists (very expensive structure) as well as a hashset. Can't we handle > this with just a simple list and a hashset? A tree of lists can scale much better and will be more efficient when dealing with a significant number of connections. However, I think it is certainly an overkill for the simple (one connection) connection manager. Mike, can we have something lighter for the SimpleHttpConnectionManager? > 3) If IdleConnectionHandler.closeIdleConnections is never called, will > the tree structure ever get cleaned up? Is there a memory leak under > this condition? Mike, is there any particular reason for not removing connections from the tree of lists in the IdleConnectionHandler#remove method? > 4) Can we add a simple utility class that implements thread and calls > IdleConnectionHandler.closeIdleConnections so that everyone doesn't have to > write their own? Agreed. A simple helper class can be quite handy Overall, the patch looks very good to me Oleg --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
