Обсуждение: read-only queries on PITRslaves, any progress?
Hello, this might not be the best place to ask, but has there been any further progress or an alternative project to Florian Pflug's SoC 2007 proposal about enabling PITR slaves to serve read-only queries? It seems like an elegant way to load-balance a PostgreSQL database with very little overhead. This was the original thread: http://archives.postgresql.org/pgsql-hackers/2007-03/msg00050.php Thanks, -mjy -- Dipl.-Ing. Marinos Yannikos, CEO Preisvergleich Internet Services AG Obere Donaustrasse 63, A-1020 Wien Tel./Fax: (+431) 5811609-52/-55 Handelsgericht Wien - FN 197241K - Firmensitz Wien
On 9/15/07, Marinos Yannikos <mjy@geizhals.at> wrote: > this might not be the best place to ask, but has there been any further > progress or an alternative project to Florian Pflug's SoC 2007 proposal > about enabling PITR slaves to serve read-only queries? It seems like an > elegant way to load-balance a PostgreSQL database with very little overhead. I too would love to hear about this. Alexander.
On 9/18/07, Alexander Staubo <alex@purefiction.net> wrote: > On 9/15/07, Marinos Yannikos <mjy@geizhals.at> wrote: > > this might not be the best place to ask, but has there been any further > > progress or an alternative project to Florian Pflug's SoC 2007 proposal > > about enabling PITR slaves to serve read-only queries? It seems like an > > elegant way to load-balance a PostgreSQL database with very little overhead. > > I too would love to hear about this. Florian has been hard at work on getting the lazy xid assignment patch in for 8.3, which is now complete. AIUI, this is an important step towards read only pitr slave (with other advantages too), and shows that things are moving along. I don't have a crystal ball, but I'm guessing you will start to hear more about this when the 8.3 beta is wrapped. merlin
On 9/18/07, Merlin Moncure <mmoncure@gmail.com> wrote: > Florian has been hard at work on getting the lazy xid assignment patch > in for 8.3, which is now complete. AIUI, this is an important step > towards read only pitr slave (with other advantages too), and shows > that things are moving along. I don't have a crystal ball, but I'm > guessing you will start to hear more about this when the 8.3 beta is > wrapped. I was hoping for a tighter timeframe, but this is good to know, thanks. Alexander.