mirror/postgresql
2
0
Fork 0
mirror of https://git.postgresql.org/git/postgresql.git synced 2025-01-24 18:55:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

pgbench: Comment on thread timing hazards.

Browse Source 
Reviewed by Fabien COELHO.
This commit is contained in:
Noah Misch 2013-10-06 09:56:43 -04:00
parent bf2617981c
commit 7db0dc0eac
1 changed files with 10 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
contrib/pgbench/pgbench.c
View File

@ -2896,7 +2896,16 @@ main(int argc, char **argv)
}
disconnect_all(state, nclients);
/* get end time */
/*
* XXX We compute results as though every client of every thread started
* and finished at the same time. That model can diverge noticeably from
* reality for a short benchmark run involving relatively many threads.
* The first thread may process notably many transactions before the last
* thread begins. Improving the model alone would bring limited benefit,
* because performance during those periods of partial thread count can
* easily exceed steady state performance. This is one of the many ways
* short runs convey deceptive performance figures.
*/
INSTR_TIME_SET_CURRENT(total_time);
INSTR_TIME_SUBTRACT(total_time, start_time);
printResults(ttype, total_xacts, nclients, threads, nthreads,