From: | Erik Jones <ejones(at)engineyard(dot)com> |
---|---|
To: | Craig Ringer <craig(at)postnewspapers(dot)com(dot)au> |
Cc: | General postgres mailing list <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Last modification time of a database? |
Date: | 2009-03-24 23:24:50 |
Message-ID: | AF2BC3CC-AE4A-4FEA-A981-139EFC02E8D9@engineyard.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Mar 23, 2009, at 5:00 PM, Craig Ringer wrote:
> Erik Jones wrote:
>
>> Am I missing something obvious here? If not, has anyone come up
>> with a reliable way to do this?
>
> Triggers on all your tables that append to a logging table?
>
> Have the client do it?
>
> Note that you do *NOT* want to have triggers that attempt to UPDATE
> a table to record the last modified time for that table. They'll
> cause transactions that touch the same table to block waiting until
> the first one commits/rolls back, so they'll ruin your concurrency.
> They may also cause unexpected deadlock aborts of transactions.
These are all client databases at the web hosting company I work at.
I can't go putting triggers on all of their tables. I think I'll just
start taking snapshots of pertinent data from pg_stat_activity and
after I've been collecting data for a while run a report of dbs that
haven't seen connections in X long since what I'm really after is
inactive databases.
Erik Jones, Database Administrator
Engine Yard
Support, Scalability, Reliability
866.518.9273 x 260
Location: US/Pacific
IRC: mage2k
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2009-03-25 00:47:39 | Re: Last modification time of a database? |
Previous Message | Brent Wood | 2009-03-24 23:22:06 | Re: Proper entry of polygon type data |