On Tue, Jul 21, 2015 at 7:20 PM, Heikki Linnakangas wrote:
> On 07/21/2015 04:14 PM, Alexander Korotkov wrote:
>
>> On Tue, Jul 21, 2015 at 3:52 PM, Heikki Linnakangas
>> wrote:
>>
>> On 07/21/2015 02:56 PM, Alexander Korotkov wrote:
>>>
>>> Probably, but currently we are in quite unlogical si
On 07/21/2015 04:14 PM, Alexander Korotkov wrote:
On Tue, Jul 21, 2015 at 3:52 PM, Heikki Linnakangas wrote:
On 07/21/2015 02:56 PM, Alexander Korotkov wrote:
Probably, but currently we are in quite unlogical situation. We have
default fillfactor = 90 for GiST where it has no use cases at al
On Tue, Jul 21, 2015 at 3:52 PM, Heikki Linnakangas wrote:
> On 07/21/2015 02:56 PM, Alexander Korotkov wrote:
>
>> Probably, but currently we are in quite unlogical situation. We have
>> default fillfactor = 90 for GiST where it has no use cases at all and
>> effectively is just a waste of space
On 07/21/2015 02:56 PM, Alexander Korotkov wrote:
Probably, but currently we are in quite unlogical situation. We have
default fillfactor = 90 for GiST where it has no use cases at all and
effectively is just a waste of space.
Why is it useless for GiST?
- Heikki
--
Sent via pgsql-hackers ma
On Tue, Jul 21, 2015 at 2:49 PM, Heikki Linnakangas wrote:
> Hmm. That's slightly different from the test case I used, in that the
> update is changing the indexed value, which means that the new value goes
> to different posting tree than the old one. I'm not sure if that makes any
> difference.
On 07/21/2015 02:00 PM, Alexander Korotkov wrote:
On Tue, Jul 21, 2015 at 12:40 PM, Heikki Linnakangas
wrote:
Has anyone done any performance testing of this?
The purpose of a fillfactor is to avoid the storm of page splits right
after building the index, when there are some random updates to
On Tue, Jul 21, 2015 at 12:40 PM, Heikki Linnakangas
wrote:
> Has anyone done any performance testing of this?
>
> The purpose of a fillfactor is to avoid the storm of page splits right
> after building the index, when there are some random updates to the table.
> It causes the index to bloat, as
Has anyone done any performance testing of this?
The purpose of a fillfactor is to avoid the storm of page splits right
after building the index, when there are some random updates to the
table. It causes the index to bloat, as every full page is split to two
half-full pages, and also adds lat
On Fri, Jul 10, 2015 at 7:13 PM, Alexander Korotkov
wrote:
> On Fri, Jul 10, 2015 at 4:54 AM, Michael Paquier <
> michael.paqu...@gmail.com> wrote:
>
>>
>>
>> On Thu, Jul 9, 2015 at 10:33 PM, Alexander Korotkov wrote:
>> > [...]
>>
>> + /* Caclculate max data size on page according to fillfac
On 07/10/2015 01:13 PM, Alexander Korotkov wrote:
On Fri, Jul 10, 2015 at 4:54 AM, Michael Paquier
wrote:
+ #define GIN_MIN_FILLFACTOR20
+ #define GIN_DEFAULT_FILLFACTOR90
I am still worrying about using a default fillfactor at 90, as we did a
lot of promotion about compress
On Fri, Jul 10, 2015 at 4:54 AM, Michael Paquier
wrote:
>
>
> On Thu, Jul 9, 2015 at 10:33 PM, Alexander Korotkov wrote:
> > [...]
>
> + /* Caclculate max data size on page according to fillfactor */
> s/Caclculate/Calculate
>
> When creating a simple gin index, I am seeing that GinGetMaxData
On Thu, Jul 9, 2015 at 10:33 PM, Alexander Korotkov wrote:
> [...]
+ /* Caclculate max data size on page according to fillfactor */
s/Caclculate/Calculate
When creating a simple gin index, I am seeing that GinGetMaxDataSize gets
called with ginEntryInsert:
* frame #0: 0x00010a49d72e
pos
Hi!
On Wed, Jul 8, 2015 at 10:27 PM, Heikki Linnakangas wrote:
> In dataPlaceToPageLeaf-function:
>
> if (append)
>> {
>> /*
>> * Even when appending, trying to append more items than
>> will fit is
>> * not completely free, beca
In dataPlaceToPageLeaf-function:
if (append)
{
/*
* Even when appending, trying to append more items than will
fit is
* not completely free, because we will merge the new items and
old
* items into an array belo
On 25.2.2015 10:20, Alexander Korotkov wrote:
> On Tue, Feb 24, 2015 at 5:15 PM, Tomas Vondra
> mailto:tomas.von...@2ndquadrant.com>> wrote:
>
>> I hoped lowering the fillfactor will improve this, but
>> fillfactor=75 had pretty much no effect in this case. Is that
>> expected for this kind of wor
On Tue, Feb 24, 2015 at 5:15 PM, Tomas Vondra
wrote:
> I hoped lowering the fillfactor will improve this, but fillfactor=75 had
> pretty much no effect in this case. Is that expected for this kind of
> workload? I see the previous discussion talked about random updates, not
> inserts, so maybe th
Hi,
I've been wondering whether this might improve behavior with one of my
workloads, suffering by GIN bloat - the same one I used to test GIN
fastscan, for example.
It's a batch process that loads a mailing list archive into a table with
a GIN index on message body, by doing something like this:
On Tue, Jan 20, 2015 at 12:06 AM, Cédric Villemain
wrote:
> Michael, I first didn't understood how GiN can benefits from the
> patch...thus my question.
>
> There were no noise for me, and I learn some more about GiN. So I thank you
> for your work!
Kicking back the ball, I haven't done as much wo
Le samedi 17 janvier 2015 08:23:44 Michael Paquier a écrit :
> On Sat, Jan 17, 2015 at 2:40 AM, Robert Haas
wrote:
> > There's only value in adding a fillfactor parameter to GIN indexes
> > if
> > it improves performance. There are no benchmarks showing it does.
> > So, why are we still talking a
On Mon, Jan 19, 2015 at 5:46 PM, Cédric Villemain
wrote:
> Le lundi 19 janvier 2015 08:24:08 Robert Haas a écrit :
>
> > On Sat, Jan 17, 2015 at 4:49 AM, Alexander Korotkov
>
> >
>
> > wrote:
>
> > > I already wrote quite detailed explanation of subject. Let mel try
>
> > > to
>
> > > explain i
Le lundi 19 janvier 2015 08:24:08 Robert Haas a écrit :
> On Sat, Jan 17, 2015 at 4:49 AM, Alexander Korotkov
>
> wrote:
> > I already wrote quite detailed explanation of subject. Let mel try
> > to
> > explain in shortly. GIN is two level nested btree. Thus, GIN would
> > have absolutely same ben
On Sat, Jan 17, 2015 at 4:49 AM, Alexander Korotkov
wrote:
> I already wrote quite detailed explanation of subject. Let mel try to
> explain in shortly. GIN is two level nested btree. Thus, GIN would have
> absolutely same benefits from fillfactor as btree. Lack of tests showing it
> is, for sure,
On Sat, Jan 17, 2015 at 12:49 PM, Alexander Korotkov
wrote:
> BTW, previous version of patch contained some bugs. Revised version is
> attached.
>
Ooops, it's here.
--
With best regards,
Alexander Korotkov.
gin_fillfactor_5.patch
Description: Binary data
--
Sent via pgsql-hackers mailin
On Fri, Jan 16, 2015 at 8:40 PM, Robert Haas wrote:
> On Thu, Jan 15, 2015 at 7:06 AM, Michael Paquier
> wrote:
> > Alexander Korotkov wrote:
> >> I'm not sure. On the one hand it's unclear why fillfactor should be
> >> different from 9.4.
> >> On the other hand it's unclear why it should be dif
On Sat, Jan 17, 2015 at 2:40 AM, Robert Haas wrote:
> There's only value in adding a fillfactor parameter to GIN indexes if
> it improves performance. There are no benchmarks showing it does.
> So, why are we still talking about this?
Indeed. There is no such benchmark as of now, and I am not sur
On Thu, Jan 15, 2015 at 7:06 AM, Michael Paquier
wrote:
> Alexander Korotkov wrote:
>> I'm not sure. On the one hand it's unclear why fillfactor should be
>> different from 9.4.
>> On the other hand it's unclear why it should be different from btree.
>> I propose marking this "ready for committer"
Alexander Korotkov wrote:
> I'm not sure. On the one hand it's unclear why fillfactor should be
> different from 9.4.
> On the other hand it's unclear why it should be different from btree.
> I propose marking this "ready for committer". So, committer can make a final
> decision.
OK let's do so the
On Thu, Jan 15, 2015 at 10:19 AM, Michael Paquier wrote:
> On Thu, Jan 8, 2015 at 2:03 PM, Michael Paquier
> wrote:
> > On Thu, Jan 8, 2015 at 6:31 AM, Alexander Korotkov
> wrote:
> >> On Wed, Jan 7, 2015 at 4:11 PM, Michael Paquier <
> michael.paqu...@gmail.com>
> >>> I am attaching an updated
On Thu, Jan 8, 2015 at 2:03 PM, Michael Paquier
wrote:
> On Thu, Jan 8, 2015 at 6:31 AM, Alexander Korotkov
> wrote:
>> On Wed, Jan 7, 2015 at 4:11 PM, Michael Paquier
>>> I am attaching an updated patch, with the default fillfactor value at
>>> 75%, and with the page split code using the fillf
On Thu, Jan 8, 2015 at 6:31 AM, Alexander Korotkov wrote:
> On Wed, Jan 7, 2015 at 4:11 PM, Michael Paquier
>> I am attaching an updated patch, with the default fillfactor value at
>> 75%, and with the page split code using the fillfactor rate.
>> Thoughts?
> Rewritten version of patch is attache
On Thu, Jan 8, 2015 at 12:31 AM, Alexander Korotkov
wrote:
>
> Rewritten version of patch is attached. I made following changes:
>
> 1) I removed fillfactor handling from entry tree. Because in this case
> fillfactor parameter would be only upper bound for actual page fullness.
> It's very like Gi
On Wed, Jan 7, 2015 at 4:11 PM, Michael Paquier
wrote:
> On Wed, Dec 3, 2014 at 2:37 AM, Robert Haas wrote:
> > On Fri, Nov 28, 2014 at 4:27 AM, Alexander Korotkov
> > wrote:
> >> On Fri, Nov 21, 2014 at 8:12 AM, Michael Paquier <
> michael.paqu...@gmail.com>
> >> wrote:
> >>> Please find attac
On Wed, Dec 3, 2014 at 2:37 AM, Robert Haas wrote:
> On Fri, Nov 28, 2014 at 4:27 AM, Alexander Korotkov
> wrote:
>> On Fri, Nov 21, 2014 at 8:12 AM, Michael Paquier
>> wrote:
>>> Please find attached a simple patch adding fillfactor as storage parameter
>>> for GIN indexes. The default value is
On Fri, Nov 28, 2014 at 4:27 AM, Alexander Korotkov
wrote:
> On Fri, Nov 21, 2014 at 8:12 AM, Michael Paquier
> wrote:
>> Please find attached a simple patch adding fillfactor as storage parameter
>> for GIN indexes. The default value is the same as the one currently aka 100
>> to have the pages
Hi!
On Fri, Nov 21, 2014 at 8:12 AM, Michael Paquier
wrote:
> Please find attached a simple patch adding fillfactor as storage parameter
> for GIN indexes. The default value is the same as the one currently aka 100
> to have the pages completely packed when a GIN index is created.
>
That's not
Le jeudi 27 novembre 2014 23:33:11 Michael Paquier a écrit :
> On Fri, Nov 21, 2014 at 2:12 PM, Michael Paquier
>
> wrote:
> > I am adding that to the commit fest of December.
>
> Here are updated patches. Alvaro notified me about an inconsistent
> comment.
what are the benefits of this patch ?
On Fri, Nov 21, 2014 at 2:12 PM, Michael Paquier
wrote:
> I am adding that to the commit fest of December.
Here are updated patches. Alvaro notified me about an inconsistent comment.
--
Michael
From eda0730d991f8b4dfbacc4d7a953ec5bff8b2ffe Mon Sep 17 00:00:00 2001
From: Michael Paquier
Date: Fri
Hi all,
Please find attached a simple patch adding fillfactor as storage parameter
for GIN indexes. The default value is the same as the one currently aka 100
to have the pages completely packed when a GIN index is created.
Note that to have this feature correctly working, the fix I sent yesterda
38 matches
Mail list logo