| From: | "Lawrence, Ramon" <ramon(dot)lawrence(at)ubc(dot)ca> |
|---|---|
| To: | "Robert Haas" <robertmhaas(at)gmail(dot)com> |
| Cc: | "Joshua Tolley" <eggyknap(at)gmail(dot)com>, "Bryce Cutt" <pandasuit(at)gmail(dot)com>, "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>, <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Proposed Patch to Improve Performance ofMulti-BatchHash Join for Skewed Data Sets |
| Date: | 2009-02-26 17:34:57 |
| Message-ID: | 6EEA43D22289484890D119821101B1DF2C199F@exchange20.mercury.ad.ubc.ca |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> That seems VERY useful - can you post the other ones (Z1, etc.) so I
> can download them all?
The Z1 data set is posted at:
http://people.ok.ubc.ca/rlawrenc/tpch1g1z.zip
I have not generated Z2, Z3, Z4 for 1G, but I can generate the Z2 and Z3
data sets, and in a hour or two they will be at:
http://people.ok.ubc.ca/rlawrenc/tpch1g2z.zip
http://people.ok.ubc.ca/rlawrenc/tpch1g3z.zip
Note that Z3 and Z4 are not really useful as the skew is extreme (98% of
the probe relation covered by top 100 values). Using the Z2/Z3 data set
should be enough to show the huge win if you do *really* have a skewed
data set.
BTW, is there any particular form/options of the pg_dump command that I
should use to make the dump?
--
Ramon Lawrence
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Robert Haas | 2009-02-26 17:46:20 | Re: Have \d show child tables that inherit from the specified parent |
| Previous Message | Robert Haas | 2009-02-26 17:25:52 | Re: Proposed Patch to Improve Performance ofMulti-BatchHash Join for Skewed Data Sets |