The virtual instance on cloud is elastic which could be increased based on need. But it should have an upper limit, right? For example, if I need an instance with 64 vCPUs and 4TB memory, and assume max physical server of the could provider is 32CPU and 2TB memory, so for this case I could not increase the virtual instance to my requirement, right?
I am thinking of this problem because I am looking for the "cloud instance elastic" meaning. If cloud virtual instance could not extend elastically unlimitedly, then when one day the work load increases 10 or 100 times, the cloud could not handle this case and I have to change the software architecture to make it scale out. That job could not be done in a short time. So the elastic on cloud is not so ideal. The upper limit could not exceed the cloud provider most powerful physical server. Is my understanding right?
For cloud providers the servers they have could only be divided, not aggregated to create a more powerful virtual instance, right?
