Hi Folks,
I'm reading up on CRR... For planned fail-over, the docs state to set the qm.ini for each QM instance on the live site, restart all the QM instances; set the qm.ini on each QM on the secondary site and restart all the QM. This seems a little slow to me and could lead to a good few minutes of down-time ?
The only way I can think of optimising it is to set up the new Live site qm.ini and restart them the second I see the original Live sites leader QM come back up - but I don't think I would know which one was the leader QM so, realistically, I'd have to wait for all three of the original live QM to come back up. I'd then restart the new live QM.
I guess I have two questions really -
1) anyone got any speedier methods?
2) (a minor issue I suspect), if i restart the original live QM one-by-one/slowly do they start to elect a leader because two of think they are still live and so I have to restart them at pretty much the same time?
Anyone got experience with CRR in a true prod environment?
thanks!
john.
------------------------------
John Hawkins
Integration Consultant
------------------------------