|
From: Prasad V. <va...@gm...> - 2013-09-19 05:49:18
|
Hi, 1) I am trying to evaluate the High Availability aspects of PGXC; and notice that GTM, and GTM-standby are configured to be in continuous sync. That means, every status change in GTM is synchronously made at GTM-standby. In such setup, what is the performance drop becoz of gtm-standby. Are there any benchmark tests run with and without GTM-standby?? what are the numbers?? 2) How is GTM failure discovered? Vanilla PGXC, doesn't integrate with clusters like Corosync, right?? 3) During GTM-failover, I see bunch of manual steps are needed to promote the GTM-standby to master; and make the GTM-proxies reconnect to the new GTM. What happens to the in-flight and new transactions while this GTM-failover happening?? I guess all active transaction will have to hang during this period, isn't?? thanks, -Prasad |
|
From: Nikhil S. <ni...@st...> - 2013-09-22 02:01:14
|
Hi Prasad, 1) I am trying to evaluate the High Availability aspects of PGXC; and > notice that GTM, and GTM-standby are configured to be in continuous > sync. That means, every status change in GTM is synchronously made at > GTM-standby. In such setup, what is the performance drop becoz of > gtm-standby. > > Are there any benchmark tests run with and without GTM-standby?? what > are the numbers?? > > We did not see any significant differences in the with and without GTM-Standby numbers when we did the runs some while ago. Don't have more specifics right now though. > 2) How is GTM failure discovered? Vanilla PGXC, doesn't integrate with > clusters like Corosync, right?? > > You can come up with your resource agents for Corosync/Pacemaker. That's what we did at StormDB. We have agents for GTM and datanode failover. > 3) During GTM-failover, I see bunch of manual steps are needed to > promote the GTM-standby to master; and make the GTM-proxies reconnect > to the new GTM. What happens to the in-flight and new transactions > while this GTM-failover happening?? > I guess all active transaction will have to hang during this period, > isn't?? > > Again if you integrate properly with Corosync/Pacemaker or have your own HA infrastructure in place, then you won't need any manual steps. Transactions would fail or error out for a brief period when this is happening. If the application has logic to retry the transactions then it might help. Regards, Nikhils > thanks, > -Prasad > > > ------------------------------------------------------------------------------ > LIMITED TIME SALE - Full Year of Microsoft Training For Just $49.99! > 1,500+ hours of tutorials including VisualStudio 2012, Windows 8, > SharePoint > 2013, SQL 2012, MVC 4, more. BEST VALUE: New Multi-Library Power Pack > includes > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9/20/13. > http://pubads.g.doubleclick.net/gampad/clk?id=58041151&iu=/4140/ostg.clktrk > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > -- StormDB - http://www.stormdb.com The Database Cloud |
|
From: Koichi S. <koi...@gm...> - 2013-09-23 01:06:17
|
GTM-master does not require sync for every command to GTM-standby. It requires sync only at the end of grouped commands from GTM-Proxy. There should not be significant influence to the performance. Regards; --- Koichi Suzuki 2013/9/22 Nikhil Sontakke <ni...@st...> > Hi Prasad, > > 1) I am trying to evaluate the High Availability aspects of PGXC; and >> notice that GTM, and GTM-standby are configured to be in continuous >> sync. That means, every status change in GTM is synchronously made at >> GTM-standby. In such setup, what is the performance drop becoz of >> gtm-standby. >> >> Are there any benchmark tests run with and without GTM-standby?? what >> are the numbers?? >> >> > We did not see any significant differences in the with and without > GTM-Standby numbers when we did the runs some while ago. Don't have more > specifics right now though. > > > >> 2) How is GTM failure discovered? Vanilla PGXC, doesn't integrate with >> clusters like Corosync, right?? >> >> > You can come up with your resource agents for Corosync/Pacemaker. That's > what we did at StormDB. We have agents for GTM and datanode failover. > > >> 3) During GTM-failover, I see bunch of manual steps are needed to >> promote the GTM-standby to master; and make the GTM-proxies reconnect >> to the new GTM. What happens to the in-flight and new transactions >> while this GTM-failover happening?? >> I guess all active transaction will have to hang during this period, >> isn't?? >> >> > Again if you integrate properly with Corosync/Pacemaker or have your own > HA infrastructure in place, then you won't need any manual steps. > > Transactions would fail or error out for a brief period when this is > happening. If the application has logic to retry the transactions then it > might help. > > Regards, > Nikhils > > >> thanks, >> -Prasad >> >> >> ------------------------------------------------------------------------------ >> LIMITED TIME SALE - Full Year of Microsoft Training For Just $49.99! >> 1,500+ hours of tutorials including VisualStudio 2012, Windows 8, >> SharePoint >> 2013, SQL 2012, MVC 4, more. BEST VALUE: New Multi-Library Power Pack >> includes >> Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9/20/13. >> >> http://pubads.g.doubleclick.net/gampad/clk?id=58041151&iu=/4140/ostg.clktrk >> _______________________________________________ >> Postgres-xc-general mailing list >> Pos...@li... >> https://lists.sourceforge.net/lists/listinfo/postgres-xc-general >> > > > > -- > StormDB - http://www.stormdb.com > The Database Cloud > > > ------------------------------------------------------------------------------ > LIMITED TIME SALE - Full Year of Microsoft Training For Just $49.99! > 1,500+ hours of tutorials including VisualStudio 2012, Windows 8, > SharePoint > 2013, SQL 2012, MVC 4, more. BEST VALUE: New Multi-Library Power Pack > includes > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9/22/13. > http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/4140/ostg.clktrk > _______________________________________________ > Postgres-xc-general mailing list > Pos...@li... > https://lists.sourceforge.net/lists/listinfo/postgres-xc-general > > |