Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.databases.postgresql > #324 > unrolled thread
| Started by | Pol <eff.off.if.you.think.youre.getting.my.email@anon.com> |
|---|---|
| First post | 2012-03-21 16:18 +0100 |
| Last post | 2012-03-23 05:04 +0000 |
| Articles | 4 — 3 participants |
Back to article view | Back to comp.databases.postgresql
Performance views/tables in PostgreSQL? Pol <eff.off.if.you.think.youre.getting.my.email@anon.com> - 2012-03-21 16:18 +0100
Re: Performance views/tables in PostgreSQL? Mladen Gogala <gogala.mladen@gmail.com> - 2012-03-22 13:16 +0000
Re: Performance views/tables in PostgreSQL? Jasen Betts <jasen@xnet.co.nz> - 2012-03-22 12:57 +0000
Re: Performance views/tables in PostgreSQL? Mladen Gogala <gogala.mladen@gmail.com> - 2012-03-23 05:04 +0000
| From | Pol <eff.off.if.you.think.youre.getting.my.email@anon.com> |
|---|---|
| Date | 2012-03-21 16:18 +0100 |
| Subject | Performance views/tables in PostgreSQL? |
| Message-ID | <4f69f15e@news.x-privat.org> |
Hi all, I'm just wondering, what is the equivalent of the Oracle Wait Interface (i.e. the V$<View_Name> system) in PostgreSQL? TIA and rgs. Paul...
[toc] | [next] | [standalone]
| From | Mladen Gogala <gogala.mladen@gmail.com> |
|---|---|
| Date | 2012-03-22 13:16 +0000 |
| Message-ID | <pan.2012.03.22.13.16.15@gmail.com> |
| In reply to | #324 |
On Wed, 21 Mar 2012 16:18:54 +0100, Pol wrote:
> Hi all,
>
> I'm just wondering, what is the equivalent of the Oracle Wait Interface
> (i.e. the V$<View_Name> system) in PostgreSQL?
>
> TIA and rgs.
>
>
> Paul...
Freeware Postgres doesn't have wait event interface. EnterpriseDB does
have one, but it isn't free.. There are some internal tables which show
statistics, the names start with "pg_".
One of the more useful ones is installed as extension:
gogala=# \d pg_stat_statements
View "public.pg_stat_statements"
Column | Type | Modifiers
---------------------+------------------+-----------
userid | oid |
dbid | oid |
query | text |
calls | bigint |
total_time | double precision |
rows | bigint |
shared_blks_hit | bigint |
shared_blks_read | bigint |
shared_blks_written | bigint |
local_blks_hit | bigint |
local_blks_read | bigint |
local_blks_written | bigint |
temp_blks_read | bigint |
temp_blks_written | bigint |
That is the basis for pg_statspack
--
http://mgogala.byethost5.com
[toc] | [prev] | [next] | [standalone]
| From | Jasen Betts <jasen@xnet.co.nz> |
|---|---|
| Date | 2012-03-22 12:57 +0000 |
| Message-ID | <jkf7kg$u20$1@reversiblemaps.ath.cx> |
| In reply to | #324 |
On 2012-03-21, Pol <eff.off.if.you.think.youre.getting.my.email@anon.com> wrote: > I'm just wondering, what is the equivalent of the Oracle > Wait Interface (i.e. the V$<View_Name> system) in PostgreSQL? explain analyze is the most commonly used query performance metric. I think there may be something else too. -- ⚂⚃ 100% natural
[toc] | [prev] | [next] | [standalone]
| From | Mladen Gogala <gogala.mladen@gmail.com> |
|---|---|
| Date | 2012-03-23 05:04 +0000 |
| Message-ID | <pan.2012.03.23.05.04.28@gmail.com> |
| In reply to | #326 |
On Thu, 22 Mar 2012 12:57:52 +0000, Jasen Betts wrote: > explain analyze is the most commonly used query performance metric. > > I think there may be something else too. With all due respect, there is nothing like the Oracle wait interface in the free version of PostgreSQL. Nada. Zilch. Explain plan answers the question "how will Postgres execute my query". The wait interface answers the question "where is the time spent". Note that explain plan answers the question about the future and wait interface answers the question about the past. Explain plan cannot tell you whether your application was waiting for lock, resolving a deadlock or simply dealing with a slow disk. It may not even be a database problem at all. Java application may have a bug ,causing it to sleep and not wake up until kissed by a prince, which doesn't happen that frequently. Oracle will tell you that it's waiting for the more data from SQL*Net, so you can start looking into the application. Postgres will also allow you to make that conclusion, but not directly. You will see no activity on the server and conclude that there is a problem with the application itself. -- http://mgogala.byethost5.com
[toc] | [prev] | [standalone]
Back to top | Article view | comp.databases.postgresql
csiph-web