[SPARK-29683] Job failed due to executor failures all available nodes are blacklisted

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 3.0.0
• Fix Version/s: 3.1.1
• Component/s: Spark Core, YARN
• Assignee: Unassigned
• Reporter: Genmao Yu
• Created: 31/Oct/19 09:35
• Updated: 01/Nov/22 03:08
• Resolved: 01/Nov/22 03:08

Description

My streaming job will fail due to executor failures all available nodes are blacklisted. This exception is thrown only when all node is blacklisted:

def isAllNodeBlacklisted: Boolean = currentBlacklistedYarnNodes.size >= numClusterNodes

val allBlacklistedNodes = excludeNodes ++ schedulerBlacklist ++ allocatorBlacklist.keySet

After diving into the code, I found some critical conditions not be handled properly:

• unchecked `excludeNodes`: it comes from user config. If not set properly, it may lead to "currentBlacklistedYarnNodes.size >= numClusterNodes". For example, we may set some nodes not in Yarn cluster.

  excludeNodes = (invalid1, invalid2, invalid3)
  clusterNodes = (valid1, valid2)
  

• `numClusterNodes` may equals 0: When HA Yarn failover, it will take some time for all NodeManagers to register ResourceManager again. In this case, `numClusterNode` may equals 0 or some other number, and Spark driver failed.
• too strong condition check: Spark driver will fail as long as "currentBlacklistedYarnNodes.size >= numClusterNodes". This condition should not indicate a unrecovered fatal. For example, there are some NodeManagers restarting. So we can give some waiting time before job failed.

Issue Links

• is caused by: SPARK-16630 Blacklist a node if executors won't launch on it.
• links to: [Github] Pull Request #28606 (cnZach)
• links to: [Github] Pull Request #35089 (sungpeo)
• links to: [Github] Pull Request #35089 (sungpeo)
• links to: GitHub Pull Request #26343