Re: WAL Sequence - Mailing list pgsql-general

From Tom Lane
Subject Re: WAL Sequence
Date
Msg-id 24439.1199735969@sss.pgh.pa.us
Whole thread Raw
In response to WAL Sequence  (Ashish Karalkar <ashish_postgre@yahoo.co.in>)
List pgsql-general
Ashish Karalkar <ashish_postgre@yahoo.co.in> writes:
> As far as my understanding every new WAL segment created is numerically greater than the previous one,but below out
showthe WAL 13 is created after wal 14,15,16,17,18,19  checked by creation time. 

You're in error to assume that the creation times are always matched to
the file names.

What is actually happening here is that the 00013 file is the latest in
actual use, and the ones that are named numerically higher are available
to use but haven't yet been written into.  Their mod times correspond to
when they were last written into under their former names as prior
segments.

            regards, tom lane

pgsql-general by date:

Previous
From: Hannes Dorbath
Date:
Subject: Re: Index trouble with 8.3b4
Next
From: "Scott Marlowe"
Date:
Subject: Re: calculating shared data memory space