[prev in list] [next in list] [prev in thread] [next in thread] 

List:       slony1-general
Subject:    Re: [Slony1-general] COPY SET vs SUBSCRIBE SET is confusing
From:       Mark Stosberg <mark () summersault ! com>
Date:       2007-01-31 14:35:37
Message-ID: epq9gg$led$1 () sea ! gmane ! org
[Download RAW message or body]

Jan Wieck wrote:
> On 1/30/2007 2:09 PM, Mark Stosberg wrote:
>> Hello,
>>
>> I suggest that these references that refer to "COPY SET" be updated to
>> refer to "SUBSCRIBE SET":
>>
>> # grep -R 'COPY SET' faq.sgml slonik_ref.sgml startslons.sgml
>> faq.sgml:Further hurting performane, the <command>COPY SET</command> event
>>
>> slonik_ref.sgml:     the <command>COPY SET</command> process winds up
>> doing more work than should be
>>
>> startslons.sgml:initial <command>COPY SET</command>.  The problem that
>> came up in that
>>
>> startslons.sgml:restarting slon, thereby restarting the <command>COPY
>> SET</command>
>>
>> startslons.sgml:<command>COPY SET</command> in progress.</para>
>>
>> ####################
>>
>> I was reading about the effects of COPY SET, and was trying to figure
>> out where the user interface to it was-- at what point in the process a
>> COPY SET happens.
>>
>> As I understand, it happens only as part of issuing a "SUBSCRIBE SET".
>> If that's correct, the documentation would improved to consistently
>> refer to operation from the perspective of the user.
>>
>> Referring to COPY SET without a connection to how it is triggered is
>> confusing.
> 
> What actually does cause the subscriber to start copying data isn't the 
> event SUBSCRIBE_SET (sent from subscriber to origin), but rather the 
> event ENABLE_SUBSCRIPTION (traveling from the sets origin to all nodes). 
> So referring everywhere to SUBSCRIBE_SET would be misleading too.

Referring to ""copy set" or "subscribe set" is like referring a
subroutine name in a user interface. It's a detail users don't need to
know about it, it seems. Refer to "Subscribe set", which is what they
interact with.

The details about "copy set" and "enable subscription" will be
interesting to some, and can be documented somewhere, even possibly
nearby, but user-centered documentation should mention /at least/ the
commands that users submit.

       Mark






_______________________________________________
Slony1-general mailing list
Slony1-general@gborg.postgresql.org
http://gborg.postgresql.org/mailman/listinfo/slony1-general
[prev in list] [next in list] [prev in thread] [next in thread]