On Wed, 2007-05-02 at 23:59 +0100, Heikki Linnakangas wrote:
Umm, you naturally have just entry per relation, but we were talking
about how many entries the table needs to hold.. You're patch had a
hard-coded value of 1000 which is quite arbitrary.
We need to think of the interaction with
Jeff Davis wrote:
On Wed, 2007-05-02 at 23:59 +0100, Heikki Linnakangas wrote:
Jeff Davis wrote:
On Wed, 2007-05-02 at 20:58 +0100, Heikki Linnakangas wrote:
Jeff Davis wrote:
What should be the maximum size of this hash table?
Good question. And also, how do you remove entries from it?
I
Simon Riggs wrote:
We need to think of the interaction with partitioning here. People will
ask whether we would recommend that individual partitions of a large
table should be larger/smaller than a particular size, to allow these
optimizations to kick in.
My thinking is that database designers
On Thu, 2007-05-03 at 09:25 +0100, Heikki Linnakangas wrote:
The hash table keeps track of ongoing seq scans. That's presumably
related to number of backends; I can't imagine a plan where a backend is
executing more than one seq scan at a time, so that gives us an upper
bound of NBackends
On Thu, 2007-05-03 at 08:01 +0100, Simon Riggs wrote:
On Wed, 2007-05-02 at 23:59 +0100, Heikki Linnakangas wrote:
Umm, you naturally have just entry per relation, but we were talking
about how many entries the table needs to hold.. You're patch had a
hard-coded value of 1000 which is
Jeff Davis wrote:
What I was trying to say before is that, in my design, it keeps track of
relations being scanned, not scans happening. So 5 backends scanning the
same table would result in one entry that consists of the most recently-
read block by any of those backends for that relation.
On Thu, 2007-05-03 at 19:27 +0100, Heikki Linnakangas wrote:
I understand that the data structure keeps track of relations being
scanned, with one entry per such relation. I think that's very good
design, and I'm not suggesting to change that.
But what's the right size for that? We don't
Hi,
I'm starting to review the synchronized scans and scan-resistant
buffer cache patches. The patches have complex interactions so I'm
taking a holistic approach.
There's four outstanding issues with the sync scans in particular:
1. The simplistic hash approach. While it's nice to not have
On Wed, 2007-05-02 at 14:26 +0100, Heikki Linnakangas wrote:
Hi,
I'm starting to review the synchronized scans and scan-resistant
buffer cache patches. The patches have complex interactions so I'm
taking a holistic approach.
There's four outstanding issues with the sync scans in
Heikki Linnakangas [EMAIL PROTECTED] writes:
Let's use a normal hash table instead, and use a lock to protect it. If we
only
update it every 10 pages or so, the overhead should be negligible. To further
reduce contention, we could modify ReadBuffer to let the caller know if the
read
Gregory Stark wrote:
Heikki Linnakangas [EMAIL PROTECTED] writes:
Let's use a normal hash table instead, and use a lock to protect it. If we only
update it every 10 pages or so, the overhead should be negligible. To further
reduce contention, we could modify ReadBuffer to let the caller know
Jeff Davis wrote:
On Wed, 2007-05-02 at 14:26 +0100, Heikki Linnakangas wrote:
Let's use a normal hash table instead, and use a lock to protect it. If
we only update it every 10 pages or so, the overhead should be
negligible. To further reduce contention, we could modify ReadBuffer to
let the
On Wed, 2007-05-02 at 20:02 +0100, Heikki Linnakangas wrote:
Gregory Stark wrote:
Heikki Linnakangas [EMAIL PROTECTED] writes:
Let's use a normal hash table instead, and use a lock to protect it. If we
only
update it every 10 pages or so, the overhead should be negligible. To
Jeff Davis wrote:
On Wed, 2007-05-02 at 20:02 +0100, Heikki Linnakangas wrote:
Gregory Stark wrote:
Heikki Linnakangas [EMAIL PROTECTED] writes:
Let's use a normal hash table instead, and use a lock to protect it. If we only
update it every 10 pages or so, the overhead should be negligible.
On Wed, 2007-05-02 at 20:58 +0100, Heikki Linnakangas wrote:
Jeff Davis wrote:
What should be the maximum size of this hash table?
Good question. And also, how do you remove entries from it?
I guess the size should somehow be related to number of backends. Each
backend will
Jeff Davis wrote:
On Wed, 2007-05-02 at 20:58 +0100, Heikki Linnakangas wrote:
Jeff Davis wrote:
What should be the maximum size of this hash table?
Good question. And also, how do you remove entries from it?
I guess the size should somehow be related to number of backends. Each
backend
On Wed, 2007-05-02 at 23:59 +0100, Heikki Linnakangas wrote:
Jeff Davis wrote:
On Wed, 2007-05-02 at 20:58 +0100, Heikki Linnakangas wrote:
Jeff Davis wrote:
What should be the maximum size of this hash table?
Good question. And also, how do you remove entries from it?
I guess the
17 matches
Mail list logo