On 12/1/14, 11:57 AM, Andres Freund wrote:
> On 2014-11-30 20:46:51 -0600, Jim Nasby wrote:
>> On 11/10/14, 7:52 PM, Tom Lane wrote:
>>> On the whole, I'm +1 for just logging the events and seeing what we learn
>>> that way. That seems like an appropriate amount of effort for finding out
>>> whether there is really an issue.
>>
>> Attached is a patch that does this.
>
> Unless somebody protests I plan to push this soon. I'll change two
> things:
>
> * Always use the same message, independent of scan_all. For one Jim's
> version would be untranslatable, for another it's not actually
> correct. In most cases we'll *not* wait, even if scan_all is
> true as we'll often just balk after !lazy_check_needs_freeze().
Good point.
> * Change the new bit in the errdetail. "could not acquire cleanup lock"
> sounds too much like an error to me. "skipped %u pinned pages" maybe?
Seems reasonable.
--
Jim Nasby, Data Architect, Blue Treble Consulting
Data in Trouble? Get it in Treble! http://BlueTreble.com