Github user uce commented on a diff in the pull request:
https://github.com/apache/incubator-flink/pull/193#discussion_r20142356
--- Diff:
flink-runtime/src/main/java/org/apache/flink/runtime/io/disk/iomanager/BlockChannelWriter.java
---
@@ -16,101 +16,40 @@
* limitations under the License.
*/
-
package org.apache.flink.runtime.io.disk.iomanager;
-
import java.io.IOException;
import java.util.concurrent.LinkedBlockingQueue;
-import java.util.concurrent.TimeUnit;
import org.apache.flink.core.memory.MemorySegment;
-
/**
* A writer that writes data in blocks to a file channel. The writer
receives the data blocks in the form of
* {@link org.apache.flink.core.memory.MemorySegment}, which it writes
entirely to the channel,
- * regardless of how space in the segment is used. The writing happens in
an asynchronous fashion. That is, a write
- * request is not processed by the thread that issues it, but by an
asynchronous writer thread. Once the request
- * is done, the asynchronous writer adds the MemorySegment to a <i>return
queue</i> where it can be popped by the
- * worker thread, to be reused. The return queue is in this case a
- * {@link java.util.concurrent.LinkedBlockingQueue}, such that the working
thread blocks until the request has been served,
- * if the request is still pending when the it requires the segment back.
- * <p>
- * Typical write behind is realized, by having a small set of segments in
the return queue at all times. When a
- * memory segment must be written, the request is issued to the writer and
a new segment is immediately popped from
- * the return queue. Once too many requests have been issued and the I/O
thread cannot keep up, the working thread
- * naturally blocks until another segment is available again.
+ * regardless of how space in the segment is used. The writing may be
realized synchronously, or asynchronously,
+ * depending on the implementation. To support
--- End diff --
trivial: comment ends apruptly
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
