Re: Fix checkpoint skip logic on idle systems by tracking LSN progress

On 11/10/16 1:03 PM, Stephen Frost wrote:
> On Thursday, November 10, 2016, Joshua D. Drake <jd@commandprompt.com
> <mailto:jd@commandprompt.com>> wrote:
> 
>     On 11/10/2016 09:33 AM, David Steele wrote:
> 
>         On 11/10/16 10:28 AM, Stephen Frost wrote:
> 
>                 diff --git a/src/backend/access/transam/xlog.c
>                 b/src/backend/access/transam/xlog.c
> 
>             [...]
> 
>                 +                       if (log_checkpoints)
>                 +                               ereport(LOG,
>                 (errmsg("checkpoint skipped")));
> 
> 
>             Do we really need to log that we're skipping a
>             checkpoint..?  As the
>             point of this is to avoid write activity on a system which
>             is idle, it
>             doesn't make sense to me to add a new cause for writes to
>             happen when
>             we're idle.
> 
> 
>         log_checkpoints is not enabled by default, though, so if the
>         user does
>         enable it don't you think they would want to know when checkpoints
>         *don't* happen?
> 
> 
>     Yes but I don't know that it needs to be anywhere below DEBUG2 (vs
>     log_checkpoints).
> 
> 
> Agreed. You certainly may wish to log checkpoints, even on an embedded
> or low I/o system, but logging that nothing is happening doesn't seem
> useful except perhaps for debugging. 

Sure, DEBUG1 or DEBUG2 makes sense.

-- 
-David
david@pgmasters.net