Re: [GENERAL] Query generator

2016-10-16 Thread Vinicius Segalin
2016-10-16 8:54 GMT-02:00 Andreas Seltenreich :

> Vinicius Segalin  writes:
>
> > 2016-09-29 16:32 GMT-03:00 Julien Rouhaud :
> >
> > >  You should try sqlsmith (https://github.com/anse1/sqlsmith), which
> works
> > >  very well.
> >
> > I had found this one before, but all I could get was queries using
> > "standard" tables, like the ones from pg_catalog and
> > information_schema. It didn't generate queries from the tables I've
> > created. Was I doing something wrong?
>
> It should pick your relations with the same likelihood it picks the
> catalog tables.  You could verify this for with, e.g.,
>
> sqlsmith --dry-run | grep your_table_name
>
> Thanks to Julien, sqlsmith's master branch now has an option
> --exclude-catalog that inhibits use of catalog relations.  If you are
> building from github, you might want to give it another chance with this
> option.  I'll probably do a proper release in next two weeks.
>
> Feedback welcome :-)
>
> regards,
> Andreas
>

Yes, I was talking in private with him and he came up with this solution.
Thank you again, Julien.
I need to try it again, did a quick test on friday, but it looks like it
works perfectly! :)


Re: [GENERAL] could not connect to server

2016-10-16 Thread Christoph Berg
Re: Antonio Silva 2016-10-16 

> 2016-10-15 06:15:20 BRT [995-1] FATAL:  data directory
> "/var/lib/postgresql/9.5/main" has group or world access
> 2016-10-15 06:15:20 BRT [995-2] DETAIL:  Permissions should be u=rwx (0700).
> 
> The problem with permissions was already indicated to  .pgpass

Hi,

upgrades do not modify data directory permissions. Something or
someone must have modified that manually. And given it also happened
to your home directory, that might have affected the whole system...

Christoph


-- 
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general


Re: [GENERAL] could not connect to server

2016-10-16 Thread Adrian Klaver

On 10/15/2016 09:09 PM, Antonio Silva wrote:

Dear Adrian and Tom, thanks for your attention.

I did an upgrade on Friday. I use to do this every week.

Log file informed:

2016-10-15 06:15:20 BRT [995-1] FATAL:  data directory
"/var/lib/postgresql/9.5/main" has group or world access
2016-10-15 06:15:20 BRT [995-2] DETAIL:  Permissions should be u=rwx (0700).

The problem with permissions was already indicated to  .pgpass

So I did a "sudo chmod 0700 main" at /var/lib/postgresql/9.5/ directory
and I could connect postgres again!

A cannot understand what happened. How an upgrade changes the
permissions of files an folders? I work with more two desktops with the
same framework (I7, 8 GB, Ubuntu 16.04, PostgreSQL, ...) and it did not
happen.


In my previous post I should have also mentioned checking syslog for the 
period in question.




Well, but I have my database available now. Thanks.

Best regards,

Antônio Olinto



--
Adrian Klaver
adrian.kla...@aklaver.com


--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general


Re: [GENERAL] could not connect to server

2016-10-16 Thread Adrian Klaver

On 10/15/2016 09:09 PM, Antonio Silva wrote:

Dear Adrian and Tom, thanks for your attention.

I did an upgrade on Friday. I use to do this every week.

Log file informed:

2016-10-15 06:15:20 BRT [995-1] FATAL:  data directory
"/var/lib/postgresql/9.5/main" has group or world access
2016-10-15 06:15:20 BRT [995-2] DETAIL:  Permissions should be u=rwx (0700).

The problem with permissions was already indicated to  .pgpass

So I did a "sudo chmod 0700 main" at /var/lib/postgresql/9.5/ directory
and I could connect postgres again!

A cannot understand what happened. How an upgrade changes the
permissions of files an folders? I work with more two desktops with the
same framework (I7, 8 GB, Ubuntu 16.04, PostgreSQL, ...) and it did not
happen.


I would look in the the /var/log/apt/history.log.* files that cover 
Friday and see what upgraded. That may offer a clue.




Well, but I have my database available now. Thanks.

Best regards,

Antônio Olinto

2016-10-15 14:28 GMT-03:00 Adrian Klaver mailto:adrian.kla...@aklaver.com>>:

On 10/15/2016 10:11 AM, Antonio Silva wrote:

Hello

I'm using postgres 9.5 on Ubuntu 16.04. I have worked with pg in my
notebook for almost a month but today postgres did not connected. I
tryed both PGAdmin and psql.


Did you do an apt-get upgrade in the interim?


I also tried to login using my own user and postgres but I
received the
messages:

Server doesn't listen
The server doesn't accept connections: the connection library
reports
could not connect to server: Conexão recusada (connection
refused) Is
the server running on host "127.0.0.1" and accepting TCP/IP
connections
on port 5432?






--
Adrian Klaver
adrian.kla...@aklaver.com


--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general


Re: [GENERAL] custom average window function failure

2016-10-16 Thread Sebastian Luque
On Sat, 15 Oct 2016 22:24:31 -0400,
Tom Lane  wrote:

> Sebastian Luque  writes:
>> Tom Lane  wrote:
>>> Are you in a position to apply patches?  It's a one-line fix:
>>> https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=dca25c2562199ce1e7e26367613912a8eadbbde8

>> I'd like to try this by obtaining the Debian source package,
>> downloading and applying patches such as this one, and then
>> rebuilding.  However, I don't know how best to download the patches
>> from the URL above.  If I click on the "patch" link, I'm simply taken
>> to the section where this is shown on the screen.  What's the
>> procecure to download these patches?

> Clicking the "patch" link and then doing "save to file" in your
> browser should produce a file that will work as a patch.

Thanks, I had completely missed the "patch" link right at the top of the
page, so was only seeing the ones below the message, which work
differently.  It all works with the top link.

-- 
Seb



-- 
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general


Re: [GENERAL] Query generator

2016-10-16 Thread Andreas Seltenreich
Vinicius Segalin  writes:

> 2016-09-29 16:32 GMT-03:00 Julien Rouhaud :
>
> >  You should try sqlsmith (https://github.com/anse1/sqlsmith), which works
> >  very well.
>
> I had found this one before, but all I could get was queries using
> "standard" tables, like the ones from pg_catalog and
> information_schema. It didn't generate queries from the tables I've
> created. Was I doing something wrong?

It should pick your relations with the same likelihood it picks the
catalog tables.  You could verify this for with, e.g.,

sqlsmith --dry-run | grep your_table_name

Thanks to Julien, sqlsmith's master branch now has an option
--exclude-catalog that inhibits use of catalog relations.  If you are
building from github, you might want to give it another chance with this
option.  I'll probably do a proper release in next two weeks.

Feedback welcome :-)

regards,
Andreas


-- 
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general


Re: [GENERAL] postgres MultiXact 9.3 corruption resolved but advice needed

2016-10-16 Thread Yvonne Murphy
Thank you kindly for your reply, 

I've read that the wraparound has to do with the amount of files in 
pg_multicast/members. So once the wraparound error happens once will it keep 
recurring  until we upgrade even though we've truncated most of the database, 

If we exported the db dropped it and then recreated it. It would create a new 
data directory with a small members directory. Would this give us some 
breathing space until we hit the corruption again

Thanks


> On 15 Oct 2016, at 20:51, Eduardo Morras  wrote:
> 
> On Sat, 15 Oct 2016 18:40:39 +
> Yvonne Murphy  wrote:
> 
>> Our application runs against postgres 9.3.9.0-2921310. We recently
>> run into the dreaded ERROR:  MultiXactId 82578299 has not been
>> created yet -- apparent wraparound. We've truncated this big table.
>> Will truncating save us from getting this error soon again or does
>> this bug have nothing to do with the size of the table? The
>> application has also been tested against. 9.3.12 will upgrading to
>> this reduce the chance of us hitting the bug until we can get the
>> application tested in 9.5. thanks
>> 
> 
> It's fixed in 9.3.14
> 
> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.postgresql.org_docs_9.3_static_release-2D9-2D3-2D14.html&d=CwIFAw&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=mQbITS1ux9CYPN1J1RXbOOYEuEJg6BkweZRAn9BwVv0&m=bqw1YuI4zfakqEw_shsiP35FhKFhlNHoDSuR_pY1vsg&s=kyUpHJqEVfkQd7AWGZDzAAJX8Z2kf2b6dzw9MWA8f3s&e=
>  
> 
> Prevent possible failure when vacuuming multixact IDs in an
> installation that has been pg_upgrade'd from pre-9.3 (Andrew Gierth,
> Álvaro Herrera) 
> The usual symptom of this bug is errors like "MultiXactId NNN has not
> been created yet -- apparent wraparound".
> 
> 
> ---   ---
> Eduardo Morras 
> 
> 
> -- 
> Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
> To make changes to your subscription:
> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.postgresql.org_mailpref_pgsql-2Dgeneral&d=CwIFAw&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=mQbITS1ux9CYPN1J1RXbOOYEuEJg6BkweZRAn9BwVv0&m=bqw1YuI4zfakqEw_shsiP35FhKFhlNHoDSuR_pY1vsg&s=B5lKCS9L9zbC5YzBmnXVfM8AtXqYF4K0i-zg-E8ADoQ&e=
>  


-- 
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general