Re: Race condition in InvalidateObsoleteReplicationSlots()

From: Álvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: Andres Freund <andres(at)anarazel(dot)de>, pgsql-hackers(at)postgresql(dot)org, Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp>, Fujii Masao <fujii(at)postgresql(dot)org>
Subject: Re: Race condition in InvalidateObsoleteReplicationSlots()
Date: 2021-06-23 14:02:12
Message-ID: 202106231402.vusydj6kl25d@alvherre.pgsql
Views: Raw Message | Whole Thread | Download mbox | Resend email
Lists: pgsql-hackers

On 2021-Jun-20, Tom Lane wrote:

> Actually ... isn't there a second race, in the opposite direction?
> IIUC, the point of this is that once we force some WAL to be sent
> to the frozen sender/receiver, they'll be killed for failure to
> respond. But the advance_wal call is not the only possible cause
> of that; a background autovacuum for example could emit some WAL.
> So I fear it's possible for the 'to release replication slot'
> message to come out before we capture $logstart. I think you
> need to capture that value before the kill not after.

I accounted for all those things and pushed again.

Álvaro Herrera Valdivia, Chile
"I can see support will not be a problem. 10 out of 10." (Simon Wittber)

In response to


Browse pgsql-hackers by date

  From Date Subject
Next Message Tomas Vondra 2021-06-23 14:14:03 Re: logical decoding and replication of sequences
Previous Message Daniel Gustafsson 2021-06-23 13:48:59 Re: Support for NSS as a libpq TLS backend