Its been a long time to blog in my blog.Hope this experience will be really help in your upcoming DC migration.
We had a customer whose DC required to migrate from the third party DC service provider to currently working service provide.We had a complication during the initial phase of the configuration due to lack network connectivity issue between DC and DR.
Customer environment which is running with couple of windows and oracle linux machines with legacy OS. During the initial phase of the migration we have issue with vsphere replication where the DC inventory will not be visible in the DR site due the operational bug with vSphere replication with has been notified to VMWare Support engineer.
So the solution to fix the issue by redeploying the vCenter and vRA(VMware Replication Applicance) in the both DC and DR.But the actual culprit is the DR site.According to VMware solution engineer the cross platform will working fine,So the troubleshooting we have recreate the DC environment(VC and vRA) in the DR site and the issue resolves.
Here comes the network,the third party DC service provider who owns the software based virtual firewall where DC power failure will be happen often,which created chaos during the replication.So what is next which think of offline data transfer and required customer to take offline clone copy of the VM and transport to our- DC.
Here is the actual part of the big headache which i’ve been tested over the weekend and followed the blog(I’ll find out the blog and update here) which is really helpful during the crisis time.
For Offline replication you need to do couple editing work in the vm configuration file.
1.You need to know the UUID of VM disk in DC which need to change the UUID in the Cloned Copy of the VM.
.Location of VM vHDD UUID mention in this article.
1.My case the lets consider the VM UUID,this is an example of the UUID.
VM1=”60 00 C2 95 78 96 04 68-53 cb d0 5e 0f a7 d0 32 .
Clone VM1=””60 00 C2 95 78 96 04 68-53 cb d0 5e 03 d6 12 .
Two ways to get the UUID of the vm vHDD
- Protect the virtual machine using vSphere replication and get the get the UUID of the disk
- Login the VM datastore in the DC location and get the UUID from vmdk filea.
2.Change the UUID in the DR site vmdk files same as DC site,Change the folder name.
3.Protect the VM using the vSphere replication and point the VM offline and UUID change vmdk to reduce the bandwidth.
Luckily customer didn’t have any windows based machines with RDM and oracle machines has the nfs volume so plan to migrate the oracle apps using the Oracle data guard replication.
Completed the Migration in 21.5 days.