Thread: To Signal The postmaster

aaliya zarrin escreveu:
> I want to signal the postmaster (User defined signal) once I created the
> trigger file (for switch over).
>  
Send a SIGHUP to postmaster. Have in mind that it doesn't work on all
supported platforms.


--  Euler Taveira de Oliveira http://www.timbira.com/

On Tue, Dec 7, 2010 at 4:22 PM, aaliya zarrin <aaliya.zarrin@gmail.com> wrote:
> Is this provision already there in the code?
> There ate already SIGUSR1 and SIGUSR2 are handled in the postmaster? is it
> for the same?

SIGUSR1 is already used. SIGUSR2 is actually unused, but reserved for children
according to the source code comment.

Though I've been thinking about this topic, a good idea doesn't occur to me yet.
http://archives.postgresql.org/pgsql-hackers/2010-09/msg00929.php

Regards,

-- 
Fujii Masao
NIPPON TELEGRAPH AND TELEPHONE CORPORATION
NTT Open Source Software Center

On 08.12.2010 06:12, aaliya zarrin wrote:
> I can not used SIGHUP, as it is already in use.
> I want to wake up the postmaster once the trigger file is generated
> and tell it to switch over from standby mode.
> Is it possible?

SIGHUP works precisely because it is already in use. When you send 
SIGHUP, the startup process will wake up immediately, and reload the 
configuration file. Reloading the configuration file is unnecessary in 
this case, but it will wake it up and it will notice the trigger file.

For 9.1, we should think of a better way to do this, perhaps using 
SIGUSR1 to wake up. Maybe we won't even need the trigger file anymore.

--   Heikki Linnakangas  EnterpriseDB   http://www.enterprisedb.com

On Wed, Dec 8, 2010 at 4:59 PM, Heikki Linnakangas
<heikki.linnakangas@enterprisedb.com> wrote:
> For 9.1, we should think of a better way to do this, perhaps using SIGUSR1
> to wake up. Maybe we won't even need the trigger file anymore.

If we use SIGUSR1, the mechanism to allow the users to specify the event type
seems to be required. For example, we should make the SIGUSR1 handler
check not only the shmem (i.e., PMSignalStat) but also the file?

Regards,

-- 
Fujii Masao
NIPPON TELEGRAPH AND TELEPHONE CORPORATION
NTT Open Source Software Center

On Wed, Dec 8, 2010 at 6:22 PM, Fujii Masao <masao.fujii@gmail.com> wrote:
> On Wed, Dec 8, 2010 at 4:59 PM, Heikki Linnakangas
> <heikki.linnakangas@enterprisedb.com> wrote:
>> For 9.1, we should think of a better way to do this, perhaps using SIGUSR1
>> to wake up. Maybe we won't even need the trigger file anymore.
>
> If we use SIGUSR1, the mechanism to allow the users to specify the event type
> seems to be required. For example, we should make the SIGUSR1 handler
> check not only the shmem (i.e., PMSignalStat) but also the file?

What I'm thinking is to make something like "pg_ctl promote" create the event
file specifying the standby promotion in $PGDATA/pg_event or elsewhere,
and send SIGUSR1 to postmaster. OTOH, when SIGUSR1 arrives, postmaster
checks whether that event file exists. If it does, postmaster removes it and
sends the signal to startup process for standby promotion.

Thought?

Regards,

-- 
Fujii Masao
NIPPON TELEGRAPH AND TELEPHONE CORPORATION
NTT Open Source Software Center

On Fri, Dec 10, 2010 at 3:51 PM, aaliya zarrin <aaliya.zarrin@gmail.com> wrote:
> Can anybody tell after finding the trigger file what steps does postgres
> follow?

The server applies all the WAL records available, performs checkpoint,
renames recovery.conf, and gets out of recovery mode.

Regards,

-- 
Fujii Masao
NIPPON TELEGRAPH AND TELEPHONE CORPORATION
NTT Open Source Software Center