Re: BUG #13188: .pgpass does not work
От | David G. Johnston |
---|---|
Тема | Re: BUG #13188: .pgpass does not work |
Дата | |
Msg-id | CAKFQuwZJwguP8wfYn0u8qxjMS3Ma5aFgLTOhtCnBc7eqyNKyOQ@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: BUG #13188: .pgpass does not work ("bzb.dev001@gmail.com" <bzb.dev001@gmail.com>) |
Ответы |
Re: BUG #13188: .pgpass does not work
|
Список | pgsql-bugs |
On Wed, Apr 29, 2015 at 11:23 AM, bzb.dev001@gmail.com <bzb.dev001@gmail.co= m > wrote: > On 2015-04-29 12:04 PM, David G. Johnston wrote: > > > On Wed, Apr 29, 2015 at 8:09 AM, bzb.dev001@gmail.com < > bzb.dev001@gmail.com> wrote: > I have even changed the ownership of the file to postgres:postgres since > I'm logged on as another user and .pgpass is located in this user's home > directory. Yet it still does not work. >> >> >> $ ls -l .pgpass >> -rw------- 1 postgres postgres 79 Apr 29 10:24 .pgpass >> > > =E2=80=8BSorry, I missed this the first time reading. This is your prob= lem. > The "postgres" Linux user and the "postgres" PostgreSQL user are not > related to each other - particularly when using md5 authentication. The > permissions on the .pgpass file must be of the user executing "psql" - > which you've indicated is not "postgres". > > So, likely your first attempt failed because you had whitespace. You > changed the permissions and now have two problems. You then fix the > whitespace and are back to a single problem. Change the permissions back > and it should now work. > > I've changed back the owner of .pgpass to the account that I've logged in > as. This is the account that I'm using to run psql. > All the spaces in .pgpass is removed. > Checked that the permission settings of .pgpass is ok. > restarted the server... > > $ sudo service postgresql restart > > Unfortunately, it is still not working. > > $ psql --host=3Dlocalhost --port=3D5432 --username=3Dpostgres --no-passw= ord > psql: fe_sendauth: no password supplied > > Incidentally, the postgresql server is running as 'postgres' user. > > =E2=80=8BI am out of ideas. I am not currently running 9.4.x but if the pr= oblem was a bug within PostgreSQL there would be more than a single report. David J.
В списке pgsql-bugs по дате отправления: