Q: 4
An architect is designing a VMware Cloud Foundation (VCF) fleet. The following information has been
provided by the customer:
Due to budget constraints, the solution must utilize the existing server hardware.
The existing server hardware consists of server models from the same vendor but different
generations.
There are ten servers available for use in this solution.
Management and Business workloads should be hosted in different clusters.
What design decision should the architect make for the lifecycle management of the solution based
on this information?
Options
Discussion
Makes sense to pick B since each cluster might have servers with different firmware and driver needs across generations. That way you avoid update failures from mismatched hardware profiles. I think that's safer, but open to counterpoints.
Option B makes the most sense. Different server generations need separate vLCM images for compatibility, so using one image for both clusters (like in D) would likely cause issues with drivers and firmware. Pretty sure that's VMware best practice.
I don’t think it’s B. D could actually work if all servers can take the same composite image-sometimes different generations from one vendor are similar enough for vLCM. Bit risky, though, since hardware quirks can cause issues. Anyone else run into this?
D if the vendor models are close enough you can usually use a single composite image.
B , separate composite images are needed because each cluster could have servers with minor differences that matter for drivers and firmware. Mixing generations in a single image (like D) usually causes compatibility headaches with vLCM. Not totally sure unless all hardware lines up perfectly, but in most real setups it's safer to split them. Anyone disagree?
Yeah, the hardware generations being different really means each cluster needs its own composite image. That's B for me. If you try to force one image across mixed gen servers, vLCM will probably choke on drivers or firmware. Someone might see it differently but that's how I've seen it play out.
B . With servers from different generations, you risk mismatched drivers if you try to standardize a single image (like D). VCF images aren't flexible enough to abstract hardware differences like that. Unless every cluster is truly identical underneath, best to do a separate composite image per cluster.
B tbh. The clusters are on different hardware generations, so a single composite image (like D) would probably cause driver conflicts. Each cluster really needs an image matched to its hardware profile. I guess if all servers had identical specs then D could make sense, but that's not the case here.
Remembered seeing almost the same scenario in a mock-definitely B. Management and workload clusters on different hardware generations just can't use a single composite image reliably.
C or B here, but pretty sure it's B. If the servers are from different generations you'll likely hit driver or firmware mismatches if you try to use a single image. Best practice is a separate vLCM image per cluster matching the underlying hardware. Makes maintenance way easier too, but open to other ideas.
Be respectful. No spam.
