[MDEV-18481] trx_rseg_update_wsrep_checkpoint() writes excessive amount of redo log Created: 2019-02-05  Updated: 2019-02-08  Resolved: 2019-02-08

Status: Closed
Project: MariaDB Server
Component/s: Galera, Storage Engine - InnoDB
Affects Version/s: 10.4.2
Fix Version/s: 10.4.3

Type: Bug Priority: Major
Reporter: Marko Mäkelä Assignee: Teemu Ollakka
Resolution: Fixed Votes: 0
Labels: performance


 Description   

The function trx_rseg_update_wsrep_checkpoint() was modified as part of the Galera 4 merge. The following code is causing warning in GCC 8.2:

	if (must_clear_rsegs) {
 
		XID null_xid;
 
		memset(&null_xid, 0, sizeof null_xid);
 
		null_xid.null();

Also, it is writing unnecessarily much redo log. 10.4 introduced the MLOG_MEMSET record, which should be used here. We should need that object null_xid at all.
Generated at Thu Feb 08 08:44:25 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.