16, 2017 at 11:47 PM, said assemlal
> wrote:
> > Just before we restart the server today, I found only one line as:
> >
> > PANIC: could not fdatasync log file 000101760083:
> Input/output
> > error
> > the database system is in recovery mode
>
&
Just before we restart the server today, I found only one line as:
PANIC: could not fdatasync log file 000101760083: Input/output
error
the database system is in recovery mode
On Mon, Oct 16, 2017 at 10:43 AM said assemlal
wrote:
> Hi,
>
> The postgresql crashed on frid
Hi,
The postgresql crashed on friday due to IO errors. It seems that the
filesystem puked.
PANIC: could not fdatasync log file 000101760077: Input/output
error
LOG: database system was interrupted; last known up at 2017-10-13 15:26:28
EDT
WARNING: terminating connection because of cra
to happen.
It is a migration case so I was thinking to take the more safe way..
So I think the behaviour is clear to me now.
Thank you.
On Mon, Nov 28, 2016 at 4:37 PM, Alvaro Herrera
wrote:
> said assemlal wrote:
> > Hello,
> >
> > PG
Hello,
PG: 9.4
CentOS 6
I am writing functions to lock results.
Let's take an example:
CREATE OR REPLACE FUNCTION usp_locking_my_result($1 VARCHAR(50), $2
VARCHAR(50))
RETURNS TEXT AS $$
SELECT value
FROM my_table
WHERE field1 = $1 AND field2 = $2 FOR UPDATE;
$$ LANGUAGE SQL;
What happens if
Hello,
I am looking for a way to test generically input arguments to raise an
exception if one is either null or empty.
I was thinking to create a function who takes an array to check them but
not sure if it's really good.
What would be the best way to achieve that ?
Thanks
Saïd
