Re: Is it ok to run vacuum full verbose command for live database for the tables which has more dead tuples?

From: Raghavendra Rao J S V <raghavendrajsv(at)gmail(dot)com>
To: Adrian Klaver <adrian(dot)klaver(at)aklaver(dot)com>
Cc: pgsql-general(at)lists(dot)postgresql(dot)org
Subject: Re: Is it ok to run vacuum full verbose command for live database for the tables which has more dead tuples?
Date: 2018-07-19 11:00:39
Message-ID: CAEHH7R6H8YH4r0XDB9GqW29+QWL0jWWb-X4rhPxCySzpEZp=QA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Please find my inline comments.

On 18 July 2018 at 18:48, Adrian Klaver <adrian(dot)klaver(at)aklaver(dot)com> wrote:

> On 07/18/2018 02:13 AM, Raghavendra Rao J S V wrote:
>
>> We have thousands of tables. But out of those tables, around 20 to 40
>> tables are always busy due to that those tables are bloating.
>>
>
> Define bloating?
>
> ​ Table contains more dead tuples more than 5000 records. Sometimes few
>> tables may contains 50million records and size also in GB's. Refer below
>> screen print.
>>
> ​

> ​
>>
>>

> In order to avoid this we are running a shell script which performs vacuum
>> full on the tables which has more than ten thousand dead tuples.
>>
>
> Out of how many live tuples?
>

​Very less ​

>
> While running this we are stopping all application processors and running
>> vacuum full on the tables which has more dead tuples.
>>
>> 1. Is it ok to run *vacuum full verbose* command for live database for
>> the tables which has more dead tuples(greater than)?
>> 2. Does it cause any *adverse *effect?
>>
>
> ​​
> https://www.postgresql.org/docs/10/static/sql-vacuum.html
>
> "FULL
>
> Selects “full” vacuum, which can reclaim more space, but takes much
> longer and exclusively locks the table. This method also requires extra
> disk space, since it writes a new copy of the table and doesn't release the
> old copy until the operation is complete. Usually this should only be used
> when a significant amount of space needs to be reclaimed from within the
> table.

​​

>
> "
>
>
>>
>> Please clarify me. Thanks in advance.
>>
>> --
>> Regards,
>> Raghavendra Rao J S V
>>
>>
>
> --
> Adrian Klaver
> adrian(dot)klaver(at)aklaver(dot)com
>

--
Regards,
Raghavendra Rao J S V
Mobile- 8861161425

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Олег Самойлов 2018-07-19 12:06:50 Incorrect description of the WITH CHECK in the row security can lead to the security issue
Previous Message Adrien NAYRAT 2018-07-19 07:35:19 Re: Shared buffers increased but cache hit ratio is still 85%