Re: pg_resetwal tests, logging, and docs update - Mailing list pgsql-hackers

From Peter Eisentraut
Subject Re: pg_resetwal tests, logging, and docs update
Date
Msg-id 41e584af-c378-6020-360d-85e51af3b417@eisentraut.org
Whole thread Raw
In response to Re: pg_resetwal tests, logging, and docs update  (Aleksander Alekseev <aleksander@timescale.com>)
List pgsql-hackers
On 13.09.23 16:36, Aleksander Alekseev wrote:
> ```
> +                // FIXME: why 2?
>                   if (set_oldest_commit_ts_xid < 2 &&
> ```
> 
> Should we rewrite this to < FrozenTransactionId ?

That's what I suspect, but we should confirm that.

> 
> ```
> +$mult = 32 * $blcksz * 4; # FIXME
> ```
> 
> Unless I'm missing something this $mult value is not used. Is it
> really needed here?

The FIXME is that I think a multiplier *should* be applied somehow.  See 
also the FIXME in the documentation for the -c option.





pgsql-hackers by date:

Previous
From: Nathan Bossart
Date:
Subject: Re: Add 'worker_type' to pg_stat_subscription
Next
From: Nathan Bossart
Date:
Subject: Re: Inefficiency in parallel pg_restore with many tables