Re: Nooby Q: Should this take five hours? And counting?

From: Kenneth Tilton <kentilton(at)gmail(dot)com>
To: Scott Marlowe <scott(dot)marlowe(at)gmail(dot)com>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: Nooby Q: Should this take five hours? And counting?
Date: 2009-04-19 00:32:39
Message-ID: 49EA7127.6010108@gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Scott Marlowe wrote:
> On Sat, Apr 18, 2009 at 4:11 PM, Kenneth Tilton <kentilton(at)gmail(dot)com>
wrote:
>> ie, 5hrs and counting, no clue how long it intends to run, but
methinks this
>> is insane even if it is 10^7 records, mebbe half a dozen dups per
value (a
>> product-id usually around 8-chars long):
>>
>> CREATE INDEX web_source_items_by_item_id_strip
>> ON web_source_items
>> USING btree (item_id_strip);
>>
>> Am I unreasonably impatient?
>>
>> I see pg getting just a few % CPU on a heavily ram/core-endowed Sun
box with
>> nothing else going on. Mebbe they installed pg on a compact flash?
DVD-RW?
>> /usr/local/something, prolly not.
>
> What does vmstat 1 60 say during the index build? Specifically the
> cpu columns for user, system, wa?

uh-oh, Unix noob too, and unfortunately someone has jumped on with a
CPU-intensive task pegging one of the cores at 100%, so these numbers
prolly do not help, but here goes:

procs -----------memory---------- ---swap-- -----io---- --system--
-----cpu------
r b swpd free buff cache si so bi bo in cs us sy
id wa st
1 1 2076312 1503204 182152 30669308 49 69 260 299 3 3
28 2 63 7 0
1 1 2076312 1502900 182152 30669656 0 0 192 2260 1198 332
25 1 50 24 0
1 1 2076312 1503024 182152 30669656 0 0 0 704 1181 282
25 1 50 25 0
1 3 2076312 1502904 182156 30669740 0 0 104 2780 1224 422
25 0 48 26 0
1 3 2076312 1502896 182156 30669740 0 0 0 1552 1173 309
25 0 50 25 0
1 1 2076312 1502140 182172 30669976 0 0 120 312 1222 396
26 1 48 25 0
1 1 2076312 1501724 182180 30670408 0 0 496 3996 1161 450
26 1 50 24 0
1 1 2076312 1501304 182200 30670960 0 0 428 2892 1203 438
26 1 50 24 0
1 2 2076312 1501064 182200 30671400 0 0 236 3456 1168 434
26 1 49 24 0
1 2 2076312 1501064 182200 30671360 0 0 0 1620 1225 357
25 0 50 25 0
1 1 2076312 1501064 182200 30671360 0 0 0 292 1205 339
25 1 49 25 0
1 1 2076312 1500652 182220 30671776 0 0 416 1740 1186 410
25 1 50 24 0
1 2 2076312 1500472 182224 30671992 0 0 208 3560 1177 399
25 1 49 24 0

I'll sample again if I get a window, but these jobs tend to run for hours.

thx,

kt

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Scott Marlowe 2009-04-19 00:55:59 Re: Nooby Q: Should this take five hours? And counting?
Previous Message Scott Marlowe 2009-04-19 00:26:04 Re: Nooby Q: Should this take five hours? And counting?