[pgpool-general: 6833] Re: Using standby pgpool instances

[Frodo Looijaard]

Hi all,

Just another small question: there is a somewhat cryptic remark in the
documentation regarding the memory_cache_enabled setting (section 5.12.1):

Note: The query cache is used the relation cache too, if
enable_shared_relcache is on. You notes that the query cache is used
regardless memory_cache_enabled parameter, when It is used by relation
cache. See Section 5.15 for more details to relation cache. 

I do not completely understand what this should mean - I think I will have
to set enabled_shared_relcache off too?

Thanks again,
  Frodo Looijaard


On Wed, 2020-01-29 at 15:45 +0100, Frodo Looijaard wrote:
> Hi all,
> 
> thanks for you clear answers. I will keep the memory query cache disabled
> (memory_cache_enabled = off).
> 
> Kind regards,
>   Frodo Looijaard
> 
>  
> On Tue, 2020-01-28 at 21:45 +0900, Tatsuo Ishii wrote:
> > Hi Frodo,
> > 
> > The only thing you need to care in this configuration is in memory
> > query cache as cache validation mechanism uses local files on each
> > pgpool node. For example if you have cache validation files for table
> > A on pgpool1, you need to invalidate (send DML command) by connecting
> > to pgpool1 aslo. But if you connect to pgpool2 and try to modify table
> > A, no cache validation will happen since the cache validation file for
> > table A is not on pgpool2.
> > 
> > Other than this, there should be no problem as Usama said.
> > 
> > Best regards,
> > --
> > Tatsuo Ishii
> > SRA OSS, Inc. Japan
> > English: http://www.sraoss.co.jp/index_en.php
> > Japanese:http://www.sraoss.co.jp
> > 
> > > Hi
> > > The quick answer is you don’t need a virtual-IP if your setup doesn’t need one, and you can connect to standby pgpool without any problem 
> > > 
> > > Thanks
> > > Best regards 
> > > Muhammad Usama 
> > > 
> > > Sent from my iPhone
> > > 
> > > > On 27-Jan-2020, at 5:33 PM, Frodo Looijaard <frodo at frodo.looijaard.name> wrote:
> > > > 
> > > > Hello,
> > > > 
> > > > I am wondering whether we need a floating IP when using multiple pgpool
> > > > instances with the watchdog (pgpool-II-4.1.0).
> > > > 
> > > > It seems you can just connect with a non-master pgpool instance.
> > > > 
> > > > Can we expect trouble when we connect to a non-master pgpool node? Can we
> > > > expect trouble when some connections use the master node and others at the
> > > > same time use a non-master node, accessing the same Postgres database?
> > > > 
> > > > If this can safely be done, we would not need a floating IP; it would
> > > > suffice to add all pgpool-nodes as alternatives to a JDBC-URL.
> > > > 
> > > > If this is not safe, would it not be a nice feature to (perhaps optionally)
> > > > disable connections to non-master pgpool nodes?
> > > > 
> > > > Thanks for your time,
> > > >  Frodo Looijaard
> > > > 
> > > > _______________________________________________
> > > > pgpool-general mailing list
> > > > pgpool-general at pgpool.net
> > > > http://www.pgpool.net/mailman/listinfo/pgpool-general
> > > _______________________________________________
> > > pgpool-general mailing list
> > > pgpool-general at pgpool.net
> > > http://www.pgpool.net/mailman/listinfo/pgpool-general
> > _______________________________________________
> > pgpool-general mailing list
> > pgpool-general at pgpool.net
> > http://www.pgpool.net/mailman/listinfo/pgpool-general
> > 
> 
> _______________________________________________
> pgpool-general mailing list
> pgpool-general at pgpool.net
> http://www.pgpool.net/mailman/listinfo/pgpool-general
>