[FLINK-9841] Web UI only show partial taskmanager log

[yanghua]

What is the purpose of the change

This pull request fixed a bug triggered web UI only show partial taskmanager log

Brief change log

• Remove the redundant resource close

Verifying this change

This change is a trivial rework / code cleanup without any test coverage.

Does this pull request potentially affect one of the following parts:

• Dependencies (does it add or upgrade a dependency): (yes / no)
• The public API, i.e., is any changed class annotated with @Public(Evolving): (yes / no)
• The serializers: (yes / no / don't know)
• The runtime per-record code paths (performance sensitive): (yes / no / don't know)
• Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Yarn/Mesos, ZooKeeper: (yes / no / don't know)
• The S3 file system connector: (yes / no / don't know)

Documentation

• Does this pull request introduce a new feature? (yes / no)
• If yes, how is the feature documented? (not applicable / docs / JavaDocs / not documented)

[zentol]

I doubt that the "double close" is causing it, a more likely scenario is that we're closing the file before it was completely written; after all Channel#writeAndFlush isn't blocking.

[yanghua]

@zentol yes, you are right, sorry about my expression. here we should not use try-with-resource, because the listener will close the file. And it seems try-with-resource closes faster than the complete listener.

[yanghua]

@dawidwys can you review this PR?

[zentol]

you could move this back into the following try block, like this:

try {
			final long fileLength = randomAccessFile.length();
			final FileChannel fileChannel = randomAccessFile.getChannel();

			try {
				[... actual handler code ...]
			} catch (Exception e) {
				fileChannel.close();
				throw e;
			}
		} catch (IOException ioe) {
			try {
				randomAccessFile.close();
			} catch (IOException e) {
				log.warn("Error while closing file.", e);
			}
		}

[yanghua]

@zentol this PR match your requirement? I hope it can be merged into 1.6, so that users can see the full taskmanager log.

[jerry-024]

@zentol yes, you are right, sorry about my expression. here we should not use try-with-resource, because the listener will close the file. And it seems try-with-resource closes faster than the complete listener.

hi @yanghua do you have test case for this solution.

[yanghua]

@jinglining Not currently, I encountered this problem in our production environment, and I fixed it and verified it. It has been merged into Flink 1.6.0, and our current production environment version is 1.6.0. It behaves normally in our environment. What problem have you encountered?