On Thu, Mar 20, 2014 at 9:17 PM, acanada wrote:
> Hello,
>
> New server postgres version is 9.3. I'm not sure if I collected latest
> statistics after migration, if you mean if the current_settings or analyze
> queries that I posted were collected after migration... yes (notice that
> there are a
Hello,
New server postgres version is 9.3. I'm not sure if I collected latest
statistics after migration, if you mean if the current_settings or analyze
queries that I posted were collected after migration... yes (notice that there
are analyze query before migration and after migration, maybe I
On Wed, Mar 19, 2014 at 10:09 PM, acanada wrote:
Hello,
>
> First of all I'd like to thank all of you for taking your time and help me
> with this. Thank you very much.
>
> I did migrate the database to the new server with 32 processors Intel(R)
> Xeon(R) CPU E5-2670 0 @ 2.60GHz and 60GB of RAM.
Hello,
new server with more ram will definitely help to keep your working set in
memory.
But if you want your queries be fast on cold (on disk) data, then you need
more/faster disks.
And work_mem = 1000MB is too much, better set to 32MB so you don’t get OOM
Killer.
And may be slightly lower sh
Hello,
I cannot do explain (analyze, buffers) since I am on 8.3 postgres version.
I am migrating to the new server and upgrading it.
Once it is ready again I will post the explain query here.
The new disk is SATA disk with 5TB, raid 0 or 1...
lspci | grep -i raid
00:1f.2 RAID bus controller: Intel
Hello Andres,
with enable_bitmapscan=off; could you do :
explain ( analyze , buffers ) select * from entity2document2 where
name='ranitidine' ;
I think it's interesting to understand how much it's clustered the table
entity2document2.
infact the query extract 13512 rows
On Tue, Mar 11, 2014 at 3:30 AM, acanada wrote:
> Hello Evgeniy!
>
> I can move the database to another server...
> This is the cat of /proc/cpuinfo. Does it have enough power or should I go
> for a better one??
>
> (It has 32 processors like this one):
>
> cat /proc/cpuinfo
> processor : 0
Hello Evgeniy!
I can move the database to another server...
This is the cat of /proc/cpuinfo. Does it have enough power or should I go for
a better one??
(It has 32 processors like this one):
cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model
On 07 Mar 2014, at 13:18, acanada wrote:
> The table entity2document2 has 30GB. In consecutive runs it gets much
> better... 30ms apron.
So you just benchmarking your hard drives with random iops.
You need more ram and faster disks.
--
Sent via pgsql-performance mailing list (pgsql-performa
El Mar 7, 2014, a las 11:03 AM, Evgeniy Shishkin escribió:
>
> On 07 Mar 2014, at 12:46, acanada wrote:
>
>>
>> El Mar 7, 2014, a las 10:39 AM, Evgeniy Shishkin escribió:
>>
>>>
Hello Mat,
Setting enable_bitmapscan to off doesn't really helps. It gets worse...
x=>
On 07 Mar 2014, at 12:46, acanada wrote:
>
> El Mar 7, 2014, a las 10:39 AM, Evgeniy Shishkin escribió:
>
>>
>>> Hello Mat,
>>>
>>> Setting enable_bitmapscan to off doesn't really helps. It gets worse...
>>>
>>> x=> SET enable_bitmapscan=off;
>>> SET
>>> x=> explain analyze select * from (
El Mar 7, 2014, a las 10:39 AM, Evgeniy Shishkin escribió:
>
>> Hello Mat,
>>
>> Setting enable_bitmapscan to off doesn't really helps. It gets worse...
>>
>> x=> SET enable_bitmapscan=off;
>> SET
>> x=> explain analyze select * from (select * from entity2document2 where
>> name='ranitidine
> Hello Mat,
>
> Setting enable_bitmapscan to off doesn't really helps. It gets worse...
>
> x=> SET enable_bitmapscan=off;
> SET
> x=> explain analyze select * from (select * from entity2document2 where
> name='ranitidine' ) as a order by a.hepval;
>
Hello Mat,
Setting enable_bitmapscan to off doesn't really helps. It gets worse...
x=> SET enable_bitmapscan=off;
SET
x=> explain analyze select * from (select * from entity2document2 where
name='ranitidine' ) as a order by a.hepval;
Il 05/mar/2014 00:36 "Venkata Balaji Nagothi" ha scritto:
>
> After looking at the distinct values, yes the composite Index on "name"
and "hepval" is not recommended. That would worsen - its expected.
>
> We need to look for other possible work around. Please drop off the above
Index. Let me see i
After looking at the distinct values, yes the composite Index on "name" and
"hepval" is not recommended. That would worsen - its expected.
We need to look for other possible work around. Please drop off the above
Index. Let me see if i can drill further into this.
Meanwhile - can you help us know
Hello,
I don't know if this helps to figure out what is the problem but after adding
the multicolumn index on name and hepval, the performance is even worse (¿?).
Ten times worse...
explain analyze select * from (select * from entity_compounddict2document
where name='progesterone') as a orde
Hello!
The table doesn't go through high inserts so I'm taking into account your
"CLUSTER" advise. Thanks.
I'm afraid that I cannot drop the indexes that don't have scans hits because
they will have scans and hits very soon
Duplicated values for this table are:
tablename | attname
On Mon, Mar 3, 2014 at 9:17 PM, acanada wrote:
> Hello,
>
> Thankyou for your answer.
> I have made more changes than a simple re-indexing recently. I have moved
> the sorting field to the table in order to avoid the join clause. Now the
> schema is very simple. The query only implies one table:
> I have simplified the query and added the last advise that you told me:
>
> Query:
>
> explain analyze select * from (select * from
entity_compounddict2document where name='ranitidine') as a order by
a.hepval;
>
Do you need full result?
If you need just top-n rows, then index on
entity_compoun
Hello,
I have changed the multicolumn index from:
"entity2document_name_qualifier_index" btree (name, qualifier)
to:
"document_qualifier_name_index" btree (qualifier, name)
And now the planner doesn't "Recheck cond:" (there are only three different
qualifiers vs. millions of n
Hello to everybody and thanks in advance to take a look to this message.
I'm new in this list and with PostgreSQL.
My queries are taking too much time to complete and I don't know what to do
right now. I think I'm providing all the info required for you to help me. If
you need extra info please
22 matches
Mail list logo
