Re: regression, deadlock in high frequency single-row UPDATE - Mailing list pgsql-bugs

From Mark Kirkwood
Subject Re: regression, deadlock in high frequency single-row UPDATE
Date
Msg-id 548CF4EF.9090104@catalyst.net.nz
Whole thread Raw
In response to Re: regression, deadlock in high frequency single-row UPDATE  (Mark Kirkwood <mark.kirkwood@catalyst.net.nz>)
List pgsql-bugs
On 14/12/14 15:22, Mark Kirkwood wrote:
> On 13/12/14 10:57, Alvaro Herrera wrote:
>> Alvaro Herrera wrote:
>>> Alvaro Herrera wrote:
>>>
>>>> I'm going to experiment with that idea and see if it leads to a
>>>> solution.  I tried the other idea yesterday (to keep the HW tuple lock
>>>> we acquire in heap_lock_tuple until heap_update is done) but aside from
>>>> being very complicated and bug-prone, it doesn't solve the problem
>>>> anyway.
>>>
>>> Here's a preliminary patch.
>>
>> Here's a finished version of this patch, which I messed a bit with and
>> so needs some extra testing.  I want to push this shortly and backpatch
>> to 9.4 and 9.3.  I don't want to disrupt the 9.4.0 release next week,
>> but it'd be nice not to ship it with this bug.
>>
>
> This patch seems good too. I'm possibly seeing it performing slightly
> slower than the previous patch (approx 68 tps vs 63 on average -
> completely untuned 9.5 running on a single SATA drive, but that could
> just be natural variation/noise).
>

Blast - and forgot to clarify that "seems good" means I've tested it
against both your and Andrew's deadlock producing schema (no deadlocks
with this patch applied)...

Cheers

Mark

pgsql-bugs by date:

Previous
From: Mark Kirkwood
Date:
Subject: Re: regression, deadlock in high frequency single-row UPDATE
Next
From: jonlachlan@gmail.com
Date:
Subject: BUG #12228: Primary use-case of PERCENT_RANK not supported