Re: InvalidXLogRecPtr in docs

On Tue, Jun 15, 2010 at 4:22 AM, Fujii Masao <masao(dot)fujii(at)gmail(dot)com> wrote:
> On Tue, Jun 15, 2010 at 2:41 PM, Heikki Linnakangas
> <heikki(dot)linnakangas(at)enterprisedb(dot)com> wrote:
>> On 15/06/10 08:23, Fujii Masao wrote:
>>>
>>> On Thu, Jun 10, 2010 at 11:06 PM, Tom Lane<tgl(at)sss(dot)pgh(dot)pa(dot)us>  wrote:
>>>>
>>>> I'm not sure if it's worth the trouble, or even a particularly smart
>>>> idea, to force the output of the status function to be monotonic
>>>> regardless of what happens underneath.  I think removing that claim
>>>> from the docs altogether is the easiest answer.
>>>
>>> We should
>>>
>>> (1) just remove "While streaming replication is in progress this will
>>>     increase monotonically." from the description about
>>> pg_last_xlog_receive_location()?
>>>
>>> or
>>>
>>> (2) add "But if streaming replication is restarted this will back off
>>>     to the beginning of current WAL file" into there?
>>>
>>> I'm for (2) since it's more informative. Thought?
>>
>> Something like (2) seems better, because even if we remove the note that it
>> increases monotonically, people might still assume that.
>
> The attached patch adds the following:
>
> -------------
> But when streaming replication is
> restarted this will back off to the replication starting position,
> which typically indicates the beginning of the WAL file including the
> record in the position which <function>pg_last_xlog_replay_location</>
> points to at the moment.
> -------------

Applied with some additional wordsmithing.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise Postgres Company