Re: AutoVacuum starvation from sinval messages

On 8 November 2012 23:58, Jeff Janes <jeff(dot)janes(at)gmail(dot)com> wrote:
> On Thu, Nov 8, 2012 at 2:50 PM, Hannu Krosing <hannu(at)2ndquadrant(dot)com> wrote:
>> On 11/08/2012 11:40 PM, Simon Riggs wrote:
>>>
>>> On 8 November 2012 20:36, Jeff Janes <jeff(dot)janes(at)gmail(dot)com> wrote:
>>>
>>>> It does not seem outrageous to me that there would be real-world
>>>> conditions in which invalidations would be sent more than once a
>>>> minute over prolonged periods, so this total starvation seems like a
>>>> bug.
>>>
>>> Yes, its a bug, but do you really believe the above? In what cases?
>
> We see lots of traffic on the mail list about people trying to dump
> several hundred thousand tables, or they can only create one database
> every two minutes, or truncating hundreds of tables at a time over and
> over again gets slow, etc. I know little about the internal of the
> invalidation code, but I would think doing that kind of thing must
> generate a lot of them.

OK, so the problem is *any* sinval. I thought you meant one sinval per
object per minute, which seemed much less likely.

I agree one sinval per minute for long periods is actually quite likely.

--
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services