From: | Steve Crawford <scrawford(at)pinpointresearch(dot)com> |
---|---|
To: | Jim Green <student(dot)northwestern(at)gmail(dot)com> |
Cc: | pgsql-general <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: huge price database question.. |
Date: | 2012-03-21 15:01:07 |
Message-ID: | 4F69ED33.3050706@pinpointresearch.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 03/20/2012 06:16 PM, Jim Green wrote:
>
> It looks like alternatives are kind of complex to me, right now my
> approach(perl dbi and prepared insert) would take about 8/9 mins to
> insert a day's data. I think I'll probably just stick with it and
> wait.
>
> the autovacuum processes does a lot of io and make my pc unusable
> while I do the data inserts.. and I tested autovacuum off with not
> much success because of they are launched due to the transaction id
> wrap around issue.
>
Something sounds wrong, here. XID wraps around somewhere around
4-billion transactions which is a substantial multiple of the entire
number of records you are trying to insert. Do you have any unusual
vacuum settings?
As mentioned by others and myself, anything that batches lots of inserts
together into a single transaction instead of each record being its own
transaction will help whether this is COPY or BEGIN, INSERT......, COMMIT.
Cheers,
Steve
From | Date | Subject | |
---|---|---|---|
Next Message | salah jubeh | 2012-03-21 15:25:30 | strange result with union |
Previous Message | David Kerr | 2012-03-21 14:53:06 | Re: postgresql.conf evaluation of duplicate keys |