[SPARK-6222] [STREAMING] All data may not be recovered from WAL when driver is killed

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 1.3.0
• Fix Version/s: 1.3.1, 1.4.0
• Component/s: DStreams
• Target Version/s: 1.3.1, 1.4.0
• Assignee: Hari Shreedharan
• Reporter: Hari Shreedharan
• Created: 09/Mar/15 08:46
• Updated: 24/Apr/15 00:36
• Resolved: 19/Mar/15 06:17

Description

When testing for our next release, our internal tests written by wypoon caught a regression in Spark Streaming between 1.2.0 and 1.3.0. The test runs FlumePolling stream to read data from Flume, then kills the Application Master. Once YARN restarts it, the test waits until no more data is to be written and verifies the original against the data on HDFS. This was passing in 1.2.0, but is failing now.

Since the test ties into Cloudera's internal infrastructure and build process, it cannot be directly run on an Apache build. But I have been working on isolating the commit that may have caused the regression. I have confirmed that it was caused by SPARK-5147 (PR # 4149). I confirmed this several times using the test and the failure is consistently reproducible.

To re-confirm, I reverted just this one commit (and Clock consolidation one to avoid conflicts), and the issue was no longer reproducible.

Since this is a data loss issue, I believe this is a blocker for Spark 1.3.0
/cc tdas, pwendell

Issue Links

• is related to: SPARK-5142 Possibly data may be ruined in Spark Streaming's WAL mechanism.
• links to: [Github] Pull Request #4964 (harishreedharan)
• links to: [Github] Pull Request #5008 (tdas)