Re: Is LISTEN/NOTIFY reliable?

<br /><br /> On 10/16/2010 09:04 AM, Robert Haas wrote: <blockquote
cite="mid:AANLkTinmP0sMVd2ErLdDba=1VftnvOcHYBE52tjxzdsB@mail.gmail.com"type="cite"><pre wrap="">On Sat, Oct 16, 2010 at
3:27AM, Tatsuo Ishii <a class="moz-txt-link-rfc2396E"
href="mailto:ishii@postgresql.org"><ishii@postgresql.org></a>wrote:
 
</pre><blockquote type="cite"><pre wrap="">Does anybody know PostgreSQL LISTEN/NOTIFY is more reliable than
previous versions?  I vaguely recall that in the previous
implementation, message sent by NOTIFY may not be reached to listner.
Does PostgreSQL 9.0's new implementation guarantee that the message is
received by the listener?
</pre></blockquote><pre wrap="">
I think it was always intended to be reliable (otherwise it's not much
good).  I think I remember a bug where notifications were being lost
on Windows under heavy load, but I thought we fixed that...
</pre></blockquote><br /> Here's what the docs say:<br /><br /><blockquote>If the same channel name is signaled
multipletimes from the same transaction with identical payload strings, the database server can decide to deliver a
singlenotification only. On the other hand, notifications with distinct payload strings will always be delivered as
distinctnotifications. Similarly, notifications from different transactions will never get folded into one
notification.Except for dropping later instances of duplicate notifications, <tt class="COMMAND">NOTIFY</tt> guarantees
thatnotifications from the same transaction get delivered in the order they were sent.<br /></blockquote><br />
cheers<br/><br /> andrew<br />