anmolanmol1234 commented on code in PR #7817:
URL: https://github.com/apache/hadoop/pull/7817#discussion_r2260188357
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/KeepAliveCache.java:
##########
@@ -259,48 +151,39 @@ public synchronized HttpClientConnection get()
* will make space for the new connection. If the cache is closed or of zero
size,
* the connection is closed and not added to the cache.
*
- * @param httpClientConnection HttpClientConnection to be cached
+ * @param conn HttpClientConnection to be cached
* @return true if the HttpClientConnection is added in active cache, false
otherwise.
*/
- public synchronized boolean put(HttpClientConnection httpClientConnection) {
- if (isClosed.get() || maxConn == 0) {
- closeHttpClientConnection(httpClientConnection);
+ public boolean add(HttpClientConnection conn) {
+ if (conn == null) {
+ LOG.warn("Attempt to add null HttpClientConnection to the cache for
account: {}",
+ accountNamePath);
return false;
}
- if (size() == maxConn) {
- closeHttpClientConnection(get(0).httpClientConnection);
- subList(0, 1).clear();
+ if (isClosed.get() || maxCacheConnections <= 0
+ || !conn.isOpen() || conn.isStale()) {
+ closeHttpClientConnection(conn);
+ return false;
+ }
+ while (size() >= maxCacheConnections) {
Review Comment:
Since this method is not synchronized two threads can enter this loop
simultaneously, both see size = maxCacheConnections, both remove one, and both
add a new one — leading to the cache momentarily holding more than
maxCacheConnections. So, the eviction logic should be synchronized
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
