We are presented with an apparently simple intervention to expand hardware to a server, adding a processor, its heatsink, an additional system fan and memory corresponding to that processor.
The server gives startup errors with the new components installed, everything and it recognizes them well. After multiple options, it was decided to leave the server as it was (the intervention was planned for noon and ended at noon).
The surprises begin and the server as it was before no longer starts, giving similar hardware errors and the individual components working.
We made the decision to move the data disks (the infrastructure is mounted on vmware) to another server that also has vmware with another PDC booted. That PDC is in slave mode.
Multiple problems begin. Like swimming against the current continuously.
1) There is no physical space on the server that starts and we want to migrate data. It's time to review everything that exists, dismantle stopped virtual machines, carefully delete data disks and finally extract disks to have free space.
2) Once we have space everything seems in order….but no. The Domain Controller is in a non-synchronizing state. We make the 'slave' domain controller active.
3) Windows Server 2012 DNS gives errors and does not start. We managed to synchronize it again after deleting it and recreating it.
4) The installation of the management program (ERP) does not work. We have to remake it again.
5) The final surprise comes. The Exchange Server starts up but once the mailboxes start synchronizing, the entire mailboxes are emptied, leaving only the downloaded content pending. It goes computer to computer, recovering mailboxes and content.
In the end the system is left stable. An intervention that was simple at noon ends at 4 in the morning. Seeing is believing. The surprises that technology can give.