Host hardware failure on hosts are quite common and every time this happens, we shut down customers' virtual servers to do migration. This is very disruptive to customer's business. Another pain point is that in instances when we need to do maintenance like the recent Specter / Meltdown, we need to do a data center wide reboot of servers.
Other cloud providers like Google cloud offer live migration which minimizes disruption to customers. In the recent Specter / Meltdown, AWS and Azure were also able to patch hosts and only require customers to reboot to patch their OS.
Customers simply can't trust critical workload to us when we lack this basic feature to assure them of maximum availability and the fact that competitors offer this makes our shortcoming more glaring.
Do not place IBM confidential, company confidential, or personal information into any field.