[DERBY-4050] Multithreaded clob update causes growth in table that does not get reclaimed

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 10.2.2.0, 10.3.3.0, 10.4.2.0, 10.5.1.1
• Fix Version/s: 10.1.3.3, 10.2.2.1, 10.3.3.1, 10.4.2.1, 10.5.1.1
• Component/s: Store
• Issue & fix info: High Value Fix, Release Note Needed
• Assignee: Katherine Marsden
• Reporter: Katherine Marsden
• Created: 06/Feb/09 00:52
• Updated: 27/Jul/11 19:57
• Resolved: 13/Feb/09 01:04

Description

Doing a multithreaded update of a Clob table causes table growth that does not get reclaimed except by compressing the table. The reproduction has a table with two threads. One thread updates row 1 repeatedly with 33,000 character clob. The other thread updates row 2 with a small clob, "hello". The problem occurs back to 10.2 but seems much worse on trunk than 10.2. The trunk database grew to 273MB on trunk after 10000 updates of each row. The 10.2 database grew only to 25MB. If the update is synchronized there is no growth.
I will attach the repro.

Issue Links

• is related to: DERBY-5356 Tracking for Derby space reclamation issues
• relates to: DERBY-4054 Multithreaded clob update with exclusive table locking causes table growth that is not reclaimed
• relates to: DERBY-4055 Space may not be reclaimed if row locks are not available after three retries
• relates to: DERBY-4056 Space is not reclaimed when attempting to insert row that violates UNIQUE constraint
• relates to: DERBY-4057 Space is not reclaimed if transaction is rolled back