| From: | "Tsunakawa, Takayuki" <tsunakawa(dot)takay(at)jp(dot)fujitsu(dot)com> |
|---|---|
| To: | "Ideriha, Takeshi" <ideriha(dot)takeshi(at)jp(dot)fujitsu(dot)com> |
| Cc: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | RE: Global shared meta cache |
| Date: | 2018-06-28 05:09:39 |
| Message-ID: | 0A3221C70F24FB45833433255569204D1FA25E63@G01JPEXMBYT05 |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
From: Ideriha, Takeshi [mailto:ideriha(dot)takeshi(at)jp(dot)fujitsu(dot)com]
> 1) Initialized with pgbench -i -s10
...
> pgbench -c48 -T60 -Msimple | 4956 |4965
> |95%
The scaling factor should be much greater than the number of clients. Otherwise, multiple clients would conflict on the same row of branches table, which might hide the additional overhead of the global metacache.
And I think -j should be 12 or so on your 16-core server, so that the pgbench client can maximize its concurrency.
Regards
Takayuki Tsunakawa
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andrey V. Lepikhov | 2018-06-28 05:36:44 | Re: [WIP] [B-Tree] Retail IndexTuple deletion |
| Previous Message | Charles Cui | 2018-06-28 05:05:47 | [GSoC] thrift compact protocol byte interface get implemented |