Re: [HACKERS] [PATCH] Logical decoding timeline following take II

2016-11-28 Thread Craig Ringer
It's unlikely this will get in as a standalone patch, so I'm closing the CF entry for it as RwF https://commitfest.postgresql.org/11/779/ It is now being tracked as part of logical decoding on standby at https://commitfest.postgresql.org/12/788/ in thread "Logical decoding on standby", which

Re: [HACKERS] [PATCH] Logical decoding timeline following take II

2016-11-15 Thread Craig Ringer
On 16 November 2016 at 12:44, Craig Ringer wrote: > Despite that, I've attached a revised version of the current approach > incorporating fixes for the issues you mentioned. It's preceded by the > patch to add an --endpos option to pg_recvlogical so that we can > properly

Re: [HACKERS] [PATCH] Logical decoding timeline following take II

2016-11-15 Thread Craig Ringer
On 12 November 2016 at 23:07, Andres Freund wrote > On 2016-10-24 17:49:13 +0200, Petr Jelinek wrote: >> + * Determine which timeline to read an xlog page from and set the >> + * XLogReaderState's state->currTLI to that timeline ID. > > "XLogReaderState's state->currTLI" - the

Re: [HACKERS] [PATCH] Logical decoding timeline following take II

2016-11-12 Thread Andres Freund
On 2016-10-24 17:49:13 +0200, Petr Jelinek wrote: > + * Determine which timeline to read an xlog page from and set the > + * XLogReaderState's state->currTLI to that timeline ID. "XLogReaderState's state->currTLI" - the state's a bit redundant. > + * The caller must also make sure it doesn't

Re: [HACKERS] [PATCH] Logical decoding timeline following take II

2016-10-24 Thread Craig Ringer
On 24 October 2016 at 23:49, Petr Jelinek wrote: > Hi Craig, > > On 01/09/16 06:08, Craig Ringer wrote: >> Hi all >> >> Attached is a rebased and updated logical decoding timeline following >> patch for 10.0. >> >> This is a pre-requisite for the pending work on logical

Re: [HACKERS] [PATCH] Logical decoding timeline following take II

2016-10-24 Thread Petr Jelinek
Hi Craig, On 01/09/16 06:08, Craig Ringer wrote: > Hi all > > Attached is a rebased and updated logical decoding timeline following > patch for 10.0. > > This is a pre-requisite for the pending work on logical decoding on > standby servers and simplified failover of logical decoding. > I went

Re: [HACKERS] [PATCH] Logical decoding timeline following take II

2016-10-02 Thread Michael Paquier
On Thu, Sep 1, 2016 at 1:08 PM, Craig Ringer wrote: > Attached is a rebased and updated logical decoding timeline following > patch for 10.0. Moved to next CF, nothing has happened since submission. -- Michael -- Sent via pgsql-hackers mailing list

[HACKERS] [PATCH] Logical decoding timeline following take II

2016-08-31 Thread Craig Ringer
Hi all Attached is a rebased and updated logical decoding timeline following patch for 10.0. This is a pre-requisite for the pending work on logical decoding on standby servers and simplified failover of logical decoding. Restating the commit message: __ Follow timeline switches in