public inbox for [email protected]
help / color / mirror / Atom feedPGConnection.getNotifications(0) doesn't block
4+ messages / 2 participants
[nested] [flat]
* PGConnection.getNotifications(0) doesn't block
@ 2023-08-03 14:04 Pawel Veselov <[email protected]>
2023-08-03 14:39 ` Re: PGConnection.getNotifications(0) doesn't block Pawel Veselov <[email protected]>
2023-08-04 08:47 ` Re: PGConnection.getNotifications(0) doesn't block rob stone <[email protected]>
0 siblings, 2 replies; 4+ messages in thread
From: Pawel Veselov @ 2023-08-03 14:04 UTC (permalink / raw)
To: pgsql-jdbc
Hello.
I'm implementing a bit that is expected to wait for published channel
notifications,
and once there are any - process them.
I'm using PGConnection.getNotifications(0), but that immediately
returns an empty
array, and doesn't block like JavaDoc says it should.
I'm on 42.6.0.
What am I missing?
Thank you.
^ permalink raw reply [nested|flat] 4+ messages in thread
* Re: PGConnection.getNotifications(0) doesn't block
2023-08-03 14:04 PGConnection.getNotifications(0) doesn't block Pawel Veselov <[email protected]>
@ 2023-08-03 14:39 ` Pawel Veselov <[email protected]>
1 sibling, 0 replies; 4+ messages in thread
From: Pawel Veselov @ 2023-08-03 14:39 UTC (permalink / raw)
To: pgsql-jdbc
Hello.
On Thu, Aug 3, 2023 at 4:04 PM Pawel Veselov <[email protected]> wrote:
>
> Hello.
>
> I'm implementing a bit that is expected to wait for published channel
> notifications,
> and once there are any - process them.
>
> I'm using PGConnection.getNotifications(0), but that immediately
> returns an empty
> array, and doesn't block like JavaDoc says it should.
>
> I'm on 42.6.0.
>
> What am I missing?
The transaction state is not IDLE, that's what's missing.
But JavaDoc doesn't mention that bit.
^ permalink raw reply [nested|flat] 4+ messages in thread
* Re: PGConnection.getNotifications(0) doesn't block
2023-08-03 14:04 PGConnection.getNotifications(0) doesn't block Pawel Veselov <[email protected]>
@ 2023-08-04 08:47 ` rob stone <[email protected]>
2023-08-04 09:49 ` Re: PGConnection.getNotifications(0) doesn't block Pawel Veselov <[email protected]>
1 sibling, 1 reply; 4+ messages in thread
From: rob stone @ 2023-08-04 08:47 UTC (permalink / raw)
To: Pawel Veselov <[email protected]>; pgsql-jdbc
Hello,
On Thu, 2023-08-03 at 16:04 +0200, Pawel Veselov wrote:
> Hello.
>
> I'm implementing a bit that is expected to wait for published channel
> notifications,
> and once there are any - process them.
>
> I'm using PGConnection.getNotifications(0), but that immediately
> returns an empty
> array, and doesn't block like JavaDoc says it should.
>
> I'm on 42.6.0.
>
> What am I missing?
>
> Thank you.
>
>
Have you read the JDBC docs
https://jdbc.postgresql.org/documentation/server-prepare
which has a section about Listen/Notify?
HTH
^ permalink raw reply [nested|flat] 4+ messages in thread
* Re: PGConnection.getNotifications(0) doesn't block
2023-08-03 14:04 PGConnection.getNotifications(0) doesn't block Pawel Veselov <[email protected]>
2023-08-04 08:47 ` Re: PGConnection.getNotifications(0) doesn't block rob stone <[email protected]>
@ 2023-08-04 09:49 ` Pawel Veselov <[email protected]>
0 siblings, 0 replies; 4+ messages in thread
From: Pawel Veselov @ 2023-08-04 09:49 UTC (permalink / raw)
To: rob stone <[email protected]>; +Cc: pgsql-jdbc
> > I'm implementing a bit that is expected to wait for published channel
> > notifications,
> > and once there are any - process them.
> >
> > I'm using PGConnection.getNotifications(0), but that immediately
> > returns an empty
> > array, and doesn't block like JavaDoc says it should.
> >
> > I'm on 42.6.0.
> >
> > What am I missing?
> >
> > Thank you.
> Have you read the JDBC docs
> https://jdbc.postgresql.org/documentation/server-prepare
> which has a section about Listen/Notify?
I've read https://access.crunchydata.com/documentation/pgjdbc/42.1.1/listennotify.html
instead. I've read the Listen/Notify section in
https://jdbc.postgresql.org/documentation/server-prepare just now,
which carries identical text. I've also read the JavaDocs for the method(s).
All these sources agree or at least don't contradict that
getNotifications(0) should block without conditions,
and none mention that the connection must be in the IDLE
transaction state.
For that matter, even
https://www.postgresql.org/docs/current/sql-listen.html doesn't say
anything about the fact that notifications are only received/can be
read if the connection is in that state. It only mentions that the
transaction that executed LISTEN must be first
committed for that connection to start seeing notifications (for
subscribed channel).
What I was effectively doing on that connection is:
BEGIN
LISTEN
COMMIT
BEGIN
UPDATE xxx;
<poll for notifications>
I've created this for tracking: https://github.com/pgjdbc/pgjdbc/issues/2937
^ permalink raw reply [nested|flat] 4+ messages in thread
end of thread, other threads:[~2023-08-04 09:49 UTC | newest]
Thread overview: 4+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2023-08-03 14:04 PGConnection.getNotifications(0) doesn't block Pawel Veselov <[email protected]>
2023-08-03 14:39 ` Pawel Veselov <[email protected]>
2023-08-04 08:47 ` rob stone <[email protected]>
2023-08-04 09:49 ` Pawel Veselov <[email protected]>
This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox