Timeline for Postgres 8.3 faster than 9.3?
Current License: CC BY-SA 3.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 3, 2014 at 18:28 | vote | accept | Thomas | ||
| Feb 3, 2014 at 18:28 | comment | added | Thomas | Well, it seems that max_connections has a more significant impact on 9.3 than 8.3. Reducing both platforms to 100 shrunk the gap between them. | |
| Jan 29, 2014 at 2:19 | comment | added | Craig Ringer | @Thomas Not intended as a "smackdown" at all. And I tend to find the minimum useful pgbench run to be around 60 mins. | |
| Jan 28, 2014 at 14:26 | comment | added | Thomas | trying w/ -T 14400, back in 8 hours :) | |
| Jan 28, 2014 at 14:21 | comment | added | Thomas | 1000 was set as a bug-fix shortly before a release, IIRC. Should have entered a pool bugfix, thanks for the smackdown, @CraigRinger :) | |
| Jan 28, 2014 at 8:16 | comment | added | Craig Ringer |
10 minutes is nowhere near long enough, you need to average out over several CHECKPOINTs. Your greater stats targets impose a cost where they're not needed. The defaults were raised because people had issues with index selection, but nothing stops you lowering them either per-table or globally. max_connections = 1000 is awful; read wiki.postgresql.org/wiki/Number_Of_Database_Connections .
|
|
| Jan 27, 2014 at 22:11 | answer | added | Pavel Stehule | timeline score: 7 | |
| Jan 27, 2014 at 20:08 | comment | added | Thomas |
Same machine, two PG installs on different ports. Dropping the default_statistics_target on 9.3 back to 10 seems to be improving the performance for the pgbench test. I guess because it's a straightforward set of queries?
|
|
| Jan 27, 2014 at 19:59 | comment | added | Philᵀᴹ | Just to clarify, was this performed on identical hardware and storage? | |
| Jan 27, 2014 at 19:04 | comment | added | Thomas | Same (default) *_costs, different (default) statistics - 10 vs 100, same (default) effective_cache_size. I may try the mailing list if I don't get much traffic here. | |
| Jan 27, 2014 at 18:57 | comment | added | user1822 |
You should post this to the Postgres performance mailing list. The general expectation is, that 9.3 should be faster than 8.3. Do both systems have the same cost (xxx_cost) and statistics (default_statistics_target) settings? What about effective_cache_size?
|
|
| Jan 27, 2014 at 17:44 | comment | added | Thomas | I know Windows is not the best platform for a database, but telling customers to setup *nix boxes is not a viable solution at this point. | |
| Jan 27, 2014 at 17:24 | history | asked | Thomas | CC BY-SA 3.0 |