Originally posted by: vicengle
I'm in the early stage of building a shadowimage based system for doing database refreshes in non-prod environments for things like user test, acceptance testing etc. The requirement is to refresh about a dozen environments from one source database. All the target systems need to get the same point in time copy of the source but they won't all refresh at the same time.
So, the plan is to creat one set of SVOLs and sync them to the source, and then create multiple copies of those SVOLs for use on the target system. In the documentation for shadowimage I believe this is called cascading. They show examples of HORCM files to perform this cascading. In the4 example the source DB luns are listed in horcm0.conf as usuam and the SVOLs primary copy is listed in horcm1.conf as usual. Then the secondary copies are in horcm0.conf and the documentation shows using "paircreate -vr" to sync the secondary SVOL groups to the primary SVOL group.
GROUP_DB1 (user test DB)
GROUP_DB2 (system test DB)
GROUP_DB3 (reporting DB)
GROUP_DB0 (SVOL copy of source) MU0
GROUP_DB1 (SVOL copy of source) MU1
GROUP_DB2 (SVOL copy of source) MU2
GROUP_DB2 (SVOL copy of source) MU3
Then sync as follows...
paircreate -vl -g GROUP_DB0
Then after splitting GROUP_DB0...
paircreate -vr -g GROUP_DB1
paircreate -vr -g GROUP_DB2
paircreate -vr -g GROUP_DB3
I believe I follow this logic but I'm wondering if anyone would be willing to share working configurations for shadowimage cascading pairs?