[HIVE-15376] Improve heartbeater scheduling for transactions

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.2.0
• Fix Version/s: 2.2.0
• Component/s: Transactions
• Target Version/s: 2.2.0
• Assignee: Wei Zheng Assign to me
• Reporter: Wei Zheng
• Created: 06/Dec/16 23:46
• Updated: 26/Jul/17 00:03
• Resolved: 22/Dec/16 19:08

Description

HIVE-12366 improved the heartbeater logic by bringing down the gap between the lock acquisition and first heartbeat, but that's not enough, there may still be some issue, e.g.
Time A: a transaction is opened
Time B: acquireLocks is called (blocking call), but it can take a long time to actually acquire the locks and return if the system is busy
Time C: as acquireLocks returns, the first heartbeat is sent
If hive.txn.timeout < C - A, then the transaction will be timed out and aborted, thus causing failure.

Issue Links

• is related to: HIVE-15500 fix the test failure dbtxnmgr_showlocks
• relates to: HIVE-12366 Refactor Heartbeater logic for transaction