|
Mailing Lists
|
Home /
Groups /
ColdFusion Server (CF-Server)
Failover servers - migrating to new datasources
I am faced with a difficult deployment situation. I have eight production CF servers running CF8. We have about 10 SQL server and Oracle database occurances supporting those CF applications.Wally Randall 08/10/10 12:56 P All the datasource information is stored in XMLBarney Boisvert 08/10/10 01:02 P We use Oracle here, and utilize the jdbc lite drivers. This allows us toDavid Hannum 08/10/10 01:19 P I am faced with a difficult deployment situation. I have eight production CF servers running CF8. We have about 10 SQL server and Oracle database occurances supporting those CF applications. Our agency is going to create a failover server system at our off-site emergency location with the following characteristics. All Database server names will be changed. All IP addresses for all servers will be changed. We have 130+ CF datasources defined on the webservers which we need to "migrate" to the failover location. My datasources all reference specific database server names and need to be modified on the target machines. Has anyone else had to deal with this problem? Any suggestions? Is there anyway to create the datasources from a source loaded at server startup? All the datasource information is stored in XML (/WEB-INF/cfusion/lib/neo-query.xml, if I recall correctly) files, and on CF8 you can just drop the XML files into another non-running server and start it up. On CF9 there is some additional security to prevent this. However, you can use the CFADMIN API to script the creation of all your datasources (and all your other admin settings). That's definitely the route I would take. Build yourself a configuration file per environment, and then use that to bootstrap your CF instance when it starts up. Works like a champ, and you can store your bootstrap file in version control to keep track of it. Plus, because it's code, you'll never have to worry about servers being configured differently for the same version of code AND you can have different server configuration move out with your code so you don't get out of sync. cheers, barneyb ----- Excess quoted text cut - see Original Post for more ----- We use Oracle here, and utilize the jdbc lite drivers. This allows us to build the failover right on the CF-Server: I'm sure it would work for other RDBMS as well . . . When you create the datasource, choose "other" Driver Class: oracle.jdbc.driver.OracleDriver JDBC.URL = jdbc:oracle:thin:@(DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = db1.your.primarly.server.url.or.ip.addy.here)(PORT = 1234)) (ADDRESS = (PROTOCOL = TCP)(HOST = db2.your.failover.server.url.or.ip.addy.here(PORT = 1234)) (LOAD_BALANCE =off) (FAILOVER = on) ) (CONNECT_DATA = (SERVICE_NAME = servicename.here) ) ) Dave Hannum On Tue, Aug 10, 2010 at 1:02 PM, Barney Boisvert <bboisvert@gmail.com>wrote: ----- Excess quoted text cut - see Original Post for more -----
|
June 18, 2013
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||