Re: Informix beats Oracle

With apologies to the IDS crowd, I respond to the following: > I do know DB2 can act on its own, having set up HADR on DB2 No generally available DB2 release yet provides "native" automatic failover with HADR. One must add on such automation if it is desired, using (typically) a cluster manager. DB2 does ship some sample scripts and has published white papers concerning integration of DB2/ HADR with cluster managers. Further automation is coming in the next release of DB2, including automatic failover without the necessity to install an additional product. Some info (and product downloads) available under open beta: http://www-306.ibm.com/software/data/db2/9/openbeta.html Regards, - Steve P. -- Steve Pearson, DB2 for Linux, UNIX, and Windows, IBM Software Group "Portland" Development Team, IBM Beaverton Lab, Beaverton, OR, USA

Steve, Thanks for the clarification. It was my understanding that if a Primary server running DB2 went offline that the Secondary would automatically fail over, as in a Primary server crashes, and the Secondary would automatically takes over. I wasn't aware that a second layer would be necessary, but then again I've not been around DB2 for a year or so, so my memory on this must be off. I hope whatever is gained in DB2 is also pushed into IDS, as this is a profound enhancement to the engine. But maybe IDS already has enough to make total fail-over of the database and applications possible with scripting? Just an FYI for folks that don't understand what this means, HP Serviceguard is around $10,000 USD per box, so a significant savings could be had if you can manage a db instance completely in the database engine, as well as other applications. Kinda like getting a DB engine and fail-over cluster software as a bonus. Thanks!! -t- Steve Pearson (news only) wrote: > With apologies to the IDS crowd, I respond to the following: > >> I do know DB2 can act on its own, having set up HADR on DB2 > > No generally available DB2 release yet provides "native" automatic > failover with HADR. One must add on such automation if it is desired, > using (typically) a cluster manager. DB2 does ship some sample > scripts and has published white papers concerning integration of DB2/ > HADR with cluster managers. > > Further automation is coming in the next release of DB2, including > automatic failover without the necessity to install an additional > product. Some info (and product downloads) available under open beta: > http://www-306.ibm.com/software/data/db2/9/openbeta.html > > Regards, > - Steve P. > -- > Steve Pearson, DB2 for Linux, UNIX, and Windows, IBM Software Group > "Portland" Development Team, IBM Beaverton Lab, Beaverton, OR, USA > -- -- -- Informix Dynamic Server Database Engine -- Proven Faster. Proven Better. -- -- http://www.informix.com --

IDS HDR has auto failover via the DRAUTO onconfig parameter: 0 signifies OFF = Do not automatically switch the server type in the HDR environment. 1 signifies RETAIN_TYPE = Automatically switch secondary to standard during an HDR failure. Switch back to secondary when restarting HDR. 2 signifies REVERSE_TYPE = Automatically switch secondary to standard on an HDR failure. Switch to primary (and switch original primary to secondary) when restarting HDR. To my knowledge this is available in 7.31, 9.3, 10.0 and 11.1 and not available in 9.4. You can have auto failover at the client level by configuring groups in your sqlhosts file if you are running a later version of the client (2.90?) g_server group - - i=1,c=0,e=server2 server1 onsoctcp server1 sql_turbo g=g_server server2 onsoctcp server2 sql_turbo g=g_server If server1 and server2 are configured for HDR you can tell your application to connect to the group (g_server) and you will be automatically connected to whatever server is the primary. If the primary fails, your application does have to be smart enough to detect the connection failure and issue a reconnect but it doesn't have to determine which instance is the primary. You should be able to test this out with the IIUG version of 10.0, the IBM trial version of 10.0 or the Cheetah beta. Andrew ----- Original Message ----- From: "Tool" <tool@thetool.com> Newsgroups: comp.databases.informix To: <informix-list@iiug.org> Sent: Monday, June 25, 2007 11:56 AM Subject: Re: Informix beats Oracle > Steve, > > Thanks for the clarification. > > It was my understanding that if a Primary server running DB2 went offline > that the Secondary would automatically fail over, as in a Primary server > crashes, and the Secondary would automatically takes over. I wasn't aware > that a second layer would be necessary, but then again I've not been > around > DB2 for a year or so, so my memory on this must be off. I hope whatever > is > gained in DB2 is also pushed into IDS, as this is a profound enhancement > to > the engine. But maybe IDS already has enough to make total fail-over of > the database and applications possible with scripting? > > Just an FYI for folks that don't understand what this means, HP > Serviceguard > is around $10,000 USD per box, so a significant savings could be had if > you > can manage a db instance completely in the database engine, as well as > other > applications. Kinda like getting a DB engine and fail-over cluster > software > as a bonus. > > Thanks!! > > -t- > > Steve Pearson (news only) wrote: >> With apologies to the IDS crowd, I respond to the following: >> >>> I do know DB2 can act on its own, having set up HADR on DB2 >> >> No generally available DB2 release yet provides "native" automatic >> failover with HADR. One must add on such automation if it is desired, >> using (typically) a cluster manager. DB2 does ship some sample >> scripts and has published white papers concerning integration of DB2/ >> HADR with cluster managers. >> >> Further automation is coming in the next release of DB2, including >> automatic failover without the necessity to install an additional >> product. Some info (and product downloads) available under open beta: >> http://www-306.ibm.com/software/data/db2/9/openbeta.html >> >> Regards, >> - Steve P. >> -- >> Steve Pearson, DB2 for Linux, UNIX, and Windows, IBM Software Group >> "Portland" Development Team, IBM Beaverton Lab, Beaverton, OR, USA >> > > -- > > -- > -- Informix Dynamic Server Database Engine -- Proven Faster. Proven > Better. -- > -- http://www.informix.com > -- > > _______________________________________________ > Informix-list mailing list > Informix-list@iiug.org > http://www.iiug.org/mailman/listinfo/informix-list

Excellent!! Thanks Andrew. Andrew Ford wrote: > IDS HDR has auto failover via the DRAUTO onconfig parameter: > > 0 signifies OFF = Do not automatically switch the server type in the HDR > environment. > > 1 signifies RETAIN_TYPE = Automatically switch secondary to standard during > an HDR failure. Switch back to secondary when restarting HDR. > > 2 signifies REVERSE_TYPE = Automatically switch secondary to standard on an > HDR failure. Switch to primary (and switch original primary to secondary) > when restarting HDR. > > To my knowledge this is available in 7.31, 9.3, 10.0 and 11.1 and not > available in 9.4. > > You can have auto failover at the client level by configuring groups in your > sqlhosts file if you are running a later version of the client (2.90?) > > g_server group - - > i=1,c=0,e=server2 > server1 onsoctcp server1 sql_turbo g=g_server > server2 onsoctcp server2 sql_turbo g=g_server > > If server1 and server2 are configured for HDR you can tell your application > to connect to the group (g_server) and you will be automatically connected > to whatever server is the primary. > > If the primary fails, your application does have to be smart enough to > detect the connection failure and issue a reconnect but it doesn't have to > determine which instance is the primary. > > You should be able to test this out with the IIUG version of 10.0, the IBM > trial version of 10.0 or the Cheetah beta. > > Andrew > > > ----- Original Message ----- > From: "Tool" <tool@thetool.com> > Newsgroups: comp.databases.informix > To: <informix-list@iiug.org> > Sent: Monday, June 25, 2007 11:56 AM > Subject: Re: Informix beats Oracle > > >> Steve, >> >> Thanks for the clarification. >> >> It was my understanding that if a Primary server running DB2 went offline >> that the Secondary would automatically fail over, as in a Primary server >> crashes, and the Secondary would automatically takes over. I wasn't aware >> that a second layer would be necessary, but then again I've not been >> around >> DB2 for a year or so, so my memory on this must be off. I hope whatever >> is >> gained in DB2 is also pushed into IDS, as this is a profound enhancement >> to >> the engine. But maybe IDS already has enough to make total fail-over of >> the database and applications possible with scripting? >> >> Just an FYI for folks that don't understand what this means, HP >> Serviceguard >> is around $10,000 USD per box, so a significant savings could be had if >> you >> can manage a db instance completely in the database engine, as well as >> other >> applications. Kinda like getting a DB engine and fail-over cluster >> software >> as a bonus. >> >> Thanks!! >> >> -t- >> >> Steve Pearson (news only) wrote: >>> With apologies to the IDS crowd, I respond to the following: >>> >>>> I do know DB2 can act on its own, having set up HADR on DB2 >>> No generally available DB2 release yet provides "native" automatic >>> failover with HADR. One must add on such automation if it is desired, >>> using (typically) a cluster manager. DB2 does ship some sample >>> scripts and has published white papers concerning integration of DB2/ >>> HADR with cluster managers. >>> >>> Further automation is coming in the next release of DB2, including >>> automatic failover without the necessity to install an additional >>> product. Some info (and product downloads) available under open beta: >>> http://www-306.ibm.com/software/data/db2/9/openbeta.html >>> >>> Regards, >>> - Steve P. >>> -- >>> Steve Pearson, DB2 for Linux, UNIX, and Windows, IBM Software Group >>> "Portland" Development Team, IBM Beaverton Lab, Beaverton, OR, USA >>> >> -- >> >> -- >> -- Informix Dynamic Server Database Engine -- Proven Faster. Proven >> Better. -- >> -- http://www.informix.com >> -- >> >> _______________________________________________ >> Informix-list mailing list >> Informix-list@iiug.org >> http://www.iiug.org/mailman/listinfo/informix-list > > -- informix ( in FOR mix ) American English. A combination of "inform" and Unix. Originally a software product running primarily on UNIX platforms, it is still widely available for UNIX, Linux, and Windows computer platforms. Informix IDS, the database engine with the future built in, is available today with current releases offering one of the fastest databases, advanced technology, and super cool technology that application developers will love. Suitable and preferable for Java and Web Applications. Easy to use, easy to admin, Informix takes up less space on your system yet delivers more power than Oracle and other enterprise database products at a fraction of the cost. Other Definitions of Informix: 1. Database software typically chosen when choosing database products that don't have a marketing program. Compares favorably with other database engines such as Oracle, but at a fraction of the cost. 2. A collection of cool, highly advanced database technologies suitable for large and small data-driven applications for Java and Web applications. Can offer more choices for complex architectures than any other database engine. 3. Easy to use, easy to admin database engine for large and small businesses and applications, requiring very little data base administration or attendance. Ideal for vertical applications requiring high availability. 4. Consistently high uptime database engine requiring little to no downtime. 5. A database engine capable of managing data at very high speeds and capacity but available at a fraction of the cost of other database engines in the same class. 6. One of the most efficient database engines used in business that delivers multiples of speed of performance of competing database products such as Oracle, at a fraction of the cost of those competing products. 7. Extremely flexible database engine with more capabilities than Oracle and other high-end database engines, but without the high cost typically associated with high-end database engines. -- -- Informix Dynamic Server Database Engine -- Proven Faster. Proven Better. -- -- http://www.informix.com --

On Jun 25, 2:47 pm, "Andrew Ford" <a...@networkip.net> wrote: > IDS HDR has auto failover via the DRAUTO onconfig parameter: > > 0 signifies OFF = Do not automatically switch the server type in the HDR > environment. > > 1 signifies RETAIN_TYPE = Automatically switch secondary to standard during > an HDR failure. Switch back to secondary when restarting HDR. > > 2 signifies REVERSE_TYPE = Automatically switch secondary to standard on an > HDR failure. Switch to primary (and switch original primary to secondary) > when restarting HDR. > > To my knowledge this is available in 7.31, 9.3, 10.0 and 11.1 and not > available in 9.4. > > You can have auto failover at the client level by configuring groups in your > sqlhosts file if you are running a later version of the client (2.90?) > > g_server group - - > i=1,c=0,e=server2 > server1 onsoctcp server1 sql_turbo g=g_server > server2 onsoctcp server2 sql_turbo g=g_server > > If server1 and server2 are configured for HDR you can tell your application > to connect to the group (g_server) and you will be automatically connected > to whatever server is the primary. > > If the primary fails, your application does have to be smart enough to > detect the connection failure and issue a reconnect but it doesn't have to > determine which instance is the primary. > > You should be able to test this out with the IIUG version of 10.0, the IBM > trial version of 10.0 or the Cheetah beta. > > Andrew > > ----- Original Message ----- > From: "Tool" <t...@thetool.com> > > Newsgroups: comp.databases.informix > To: <informix-l...@iiug.org> > Sent: Monday, June 25, 2007 11:56 AM > Subject: Re: Informix beats Oracle > > > Steve, > > > Thanks for the clarification. > > > It was my understanding that if a Primary server running DB2 went offline > > that the Secondary would automatically fail over, as in a Primary server > > crashes, and the Secondary would automatically takes over. I wasn't aware > > that a second layer would be necessary, but then again I've not been > > around > > DB2 for a year or so, so my memory on this must be off. I hope whatever > > is > > gained in DB2 is also pushed into IDS, as this is a profound enhancement > > to > > the engine. But maybe IDS already has enough to make total fail-over of > > the database and applications possible with scripting? > > > Just an FYI for folks that don't understand what this means, HP > > Serviceguard > > is around $10,000 USD per box, so a significant savings could be had if > > you > > can manage a db instance completely in the database engine, as well as > > other > > applications. Kinda like getting a DB engine and fail-over cluster > > software > > as a bonus. > > > Thanks!! > > > -t- > > > Steve Pearson (news only) wrote: > >> With apologies to the IDS crowd, I respond to the following: > > >>> I do know DB2 can act on its own, having set up HADR on DB2 > > >> No generally available DB2 release yet provides "native" automatic > >> failover with HADR. One must add on such automation if it is desired, > >> using (typically) a cluster manager. DB2 does ship some sample > >> scripts and has published white papers concerning integration of DB2/ > >> HADR with cluster managers. > > >> Further automation is coming in the next release of DB2, including > >> automatic failover without the necessity to install an additional > >> product. Some info (and product downloads) available under open beta: > >>http://www-306.ibm.com/software/data/db2/9/openbeta.html > > >> Regards, > >> - Steve P. > >> -- > >> Steve Pearson, DB2 for Linux, UNIX, and Windows, IBM Software Group > >> "Portland" Development Team, IBM Beaverton Lab, Beaverton, OR, USA > > > -- > > > -- > > -- Informix Dynamic Server Database Engine -- Proven Faster. Proven > > Better. -- > > --http://www.informix.com > > -- > > > _______________________________________________ > > Informix-list mailing list > > Informix-l...@iiug.org > >http://www.iiug.org/mailman/listinfo/informix-list BUT I don't know of anyone in the right mind that would use DRAUTO for anything important. What do you think will happen if the network connection from the primary and secondary breaks? What do you think will happen once the network comes back up? You will end up with both database servers as PRIMARY. This is why Microsoft implemented database mirroring with automatic fail-over using a third 'witness' server (SQL Server 2005). With a quorum of at least two servers you can guarantee that this condition would not occur. The third server doesn't need to be anything fancy, it just checks to see if either the primary or the secondary can be reached. And you don't need the full enterprise license of this third server.

> On Jun 25, 2:47 pm, "Andrew Ford" <a...@networkip.net> wrote: >> IDS HDR has auto failover via the DRAUTO onconfig parameter: >> >> 0 signifies OFF = Do not automatically switch the server type in the HDR >> environment. >> >> 1 signifies RETAIN_TYPE = Automatically switch secondary to standard >> during >> an HDR failure. Switch back to secondary when restarting HDR. >> >> 2 signifies REVERSE_TYPE = Automatically switch secondary to standard on >> an >> HDR failure. Switch to primary (and switch original primary to secondary) >> when restarting HDR. >> >> To my knowledge this is available in 7.31, 9.3, 10.0 and 11.1 and not >> available in 9.4. >> >> You can have auto failover at the client level by configuring groups in >> your >> sqlhosts file if you are running a later version of the client (2.90?) >> >> g_server group - - >> i=1,c=0,e=server2 >> server1 onsoctcp server1 sql_turbo g=g_server >> server2 onsoctcp server2 sql_turbo g=g_server >> >> If server1 and server2 are configured for HDR you can tell your >> application >> to connect to the group (g_server) and you will be automatically >> connected >> to whatever server is the primary. >> >> If the primary fails, your application does have to be smart enough to >> detect the connection failure and issue a reconnect but it doesn't have >> to >> determine which instance is the primary. >> >> You should be able to test this out with the IIUG version of 10.0, the >> IBM >> trial version of 10.0 or the Cheetah beta. >> >> Andrew >> >> ----- Original Message ----- >> From: "Tool" <t...@thetool.com> >> >> Newsgroups: comp.databases.informix >> To: <informix-l...@iiug.org> >> Sent: Monday, June 25, 2007 11:56 AM >> Subject: Re: Informix beats Oracle >> >> > Steve, >> >> > Thanks for the clarification. >> >> > It was my understanding that if a Primary server running DB2 went >> > offline >> > that the Secondary would automatically fail over, as in a Primary >> > server >> > crashes, and the Secondary would automatically takes over. I wasn't >> > aware >> > that a second layer would be necessary, but then again I've not been >> > around >> > DB2 for a year or so, so my memory on this must be off. I hope >> > whatever >> > is >> > gained in DB2 is also pushed into IDS, as this is a profound >> > enhancement >> > to >> > the engine. But maybe IDS already has enough to make total fail-over >> > of >> > the database and applications possible with scripting? >> >> > Just an FYI for folks that don't understand what this means, HP >> > Serviceguard >> > is around $10,000 USD per box, so a significant savings could be had if >> > you >> > can manage a db instance completely in the database engine, as well as >> > other >> > applications. Kinda like getting a DB engine and fail-over cluster >> > software >> > as a bonus. >> >> > Thanks!! >> >> > -t- >> >> > Steve Pearson (news only) wrote: >> >> With apologies to the IDS crowd, I respond to the following: >> >> >>> I do know DB2 can act on its own, having set up HADR on DB2 >> >> >> No generally available DB2 release yet provides "native" automatic >> >> failover with HADR. One must add on such automation if it is desired, >> >> using (typically) a cluster manager. DB2 does ship some sample >> >> scripts and has published white papers concerning integration of DB2/ >> >> HADR with cluster managers. >> >> >> Further automation is coming in the next release of DB2, including >> >> automatic failover without the necessity to install an additional >> >> product. Some info (and product downloads) available under open beta: >> >>http://www-306.ibm.com/software/data/db2/9/openbeta.html >> >> >> Regards, >> >> - Steve P. >> >> -- >> >> Steve Pearson, DB2 for Linux, UNIX, and Windows, IBM Software Group >> >> "Portland" Development Team, IBM Beaverton Lab, Beaverton, OR, USA >> >> > -- >> >> > -- >> > -- Informix Dynamic Server Database Engine -- Proven Faster. Proven >> > Better. -- >> > --http://www.informix.com >> > -- >> >> > _______________________________________________ >> > Informix-list mailing list >> > Informix-l...@iiug.org >> >http://www.iiug.org/mailman/listinfo/informix-list > > BUT I don't know of anyone in the right mind that would use DRAUTO for > anything important. What do you think will happen if the network > connection from the primary and secondary breaks? What do you think > will happen once the network comes back up? > > You will end up with both database servers as PRIMARY. This is why > Microsoft implemented database mirroring with automatic fail-over > using a third 'witness' server (SQL Server 2005). With a quorum of at > least two servers you can guarantee that this condition would not > occur. The third server doesn't need to be anything fancy, it just > checks to see if either the primary or the secondary can be reached. > And you don't need the full enterprise license of this third server. I believe this is reason the feature was removed from 9.4, kind of more tech support trouble than it was worth. But if you have a reliable network (cross connect going from NIC to NIC and no hubs or networking equipment in the middle is a good start) this can be usefull. If your chances of network failure are less than the chances of hardware or os failure (IDS doesn't fail, right) then this feature has benefit.

> -----Original Message----- > From: informix-list-bounces@iiug.org > [mailto:informix-list-bounces@iiug.org] On Behalf Of Jim Kenedy > Sent: Monday, June 25, 2007 1:38 PM > To: informix-list@iiug.org > Subject: Re: IDS auto failover (was Informix beats Oracle) > > On Jun 25, 2:47 pm, "Andrew Ford" <a...@networkip.net> wrote: > > IDS HDR has auto failover via the DRAUTO onconfig parameter: > > > > 0 signifies OFF = Do not automatically switch the server > type in the HDR > > environment. > > > > 1 signifies RETAIN_TYPE = Automatically switch secondary to > standard during > > an HDR failure. Switch back to secondary when restarting HDR. > > > > 2 signifies REVERSE_TYPE = Automatically switch secondary > to standard on an > > HDR failure. Switch to primary (and switch original primary > to secondary) > > when restarting HDR. > > > > To my knowledge this is available in 7.31, 9.3, 10.0 and > 11.1 and not > > available in 9.4. > > > > You can have auto failover at the client level by > configuring groups in your > > sqlhosts file if you are running a later version of the > client (2.90?) > > > > g_server group - - > > i=1,c=0,e=server2 > > server1 onsoctcp server1 sql_turbo > g=g_server > > server2 onsoctcp server2 sql_turbo > g=g_server> > > > If server1 and server2 are configured for HDR you can tell > your application > > to connect to the group (g_server) and you will be > automatically connected > > to whatever server is the primary. > > > > If the primary fails, your application does have to be > smart enough to > > detect the connection failure and issue a reconnect but it > doesn't have to > > determine which instance is the primary. > > > > You should be able to test this out with the IIUG version > of 10.0, the IBM > > trial version of 10.0 or the Cheetah beta. > > > > Andrew > > > > ----- Original Message ----- > > From: "Tool" <t...@thetool.com> > > > > Newsgroups: comp.databases.informix > > To: <informix-l...@iiug.org> > > Sent: Monday, June 25, 2007 11:56 AM > > Subject: Re: Informix beats Oracle > > > > > Steve, > > > > > Thanks for the clarification. > > > > > It was my understanding that if a Primary server running > DB2 went offline > > > that the Secondary would automatically fail over, as in a > Primary server > > > crashes, and the Secondary would automatically takes > over. I wasn't aware > > > that a second layer would be necessary, but then again > I've not been > > > around > > > DB2 for a year or so, so my memory on this must be off. > I hope whatever > > > is > > > gained in DB2 is also pushed into IDS, as this is a > profound enhancement > > > to > > > the engine. But maybe IDS already has enough to make > total fail-over of > > > the database and applications possible with scripting? > > > > > Just an FYI for folks that don't understand what this means, HP > > > Serviceguard > > > is around $10,000 USD per box, so a significant savings > could be had if > > > you > > > can manage a db instance completely in the database > engine, as well as > > > other > > > applications. Kinda like getting a DB engine and > fail-over cluster > > > software > > > as a bonus. > > > > > Thanks!! > > > > > -t- > > > > > Steve Pearson (news only) wrote: > > >> With apologies to the IDS crowd, I respond to the following: > > > > >>> I do know DB2 can act on its own, having set up HADR on DB2 > > > > >> No generally available DB2 release yet provides "native" > automatic > > >> failover with HADR. One must add on such automation if > it is desired, > > >> using (typically) a cluster manager. DB2 does ship some sample > > >> scripts and has published white papers concerning > integration of DB2/ > > >> HADR with cluster managers. > > > > >> Further automation is coming in the next release of DB2, > including > > >> automatic failover without the necessity to install an additional > > >> product. Some info (and product downloads) available > under open beta: > > >>http://www-306.ibm.com/software/data/db2/9/openbeta.html > > > > >> Regards, > > >> - Steve P. > > >> -- > > >> Steve Pearson, DB2 for Linux, UNIX, and Windows, IBM > Software Group > > >> "Portland" Development Team, IBM Beaverton Lab, > Beaverton, OR, USA > > > > > -- > > > > > -- > > > -- Informix Dynamic Server Database Engine -- Proven > Faster. Proven > > > Better. -- > > > --http://www.informix.com > > > -- > > > > > _______________________________________________ > > > Informix-list mailing list > > > Informix-l...@iiug.org > > >http://www.iiug.org/mailman/listinfo/informix-list > > BUT I don't know of anyone in the right mind that would use DRAUTO for > anything important. What do you think will happen if the network > connection from the primary and secondary breaks? What do you think > will happen once the network comes back up? > > You will end up with both database servers as PRIMARY. This is why > Microsoft implemented database mirroring with automatic fail-over > using a third 'witness' server (SQL Server 2005). With a quorum of at > least two servers you can guarantee that this condition would not > occur. The third server doesn't need to be anything fancy, it just > checks to see if either the primary or the secondary can be reached. > And you don't need the full enterprise license of this third server. > We did a home-grown version of that for our highly-critical IDS HDR pair. http://www.iiug.org/idug06/M10.pdf HTH, Paul M.