VMware
Question 1: Thick Provisioning Types
Lazy Zeroed: Allocates all space at creation but zeros out blocks only when they are first written. Faster to create, slightly slower on first write.
Eager Zeroed: Allocates and zeros out all disk blocks at creation. Slower to create but best performance during runtime.
Question 2: Memory Reclaiming Process
Transparent Page Sharing (TPS): Eliminates redundant memory pages across VMs.
Memory Ballooning: Uses VMware tools driver to reclaim unused memory from VMs.
Memory Compression: Compresses memory pages before swapping to disk to reduce latency.
Memory Swap: Swaps VM memory pages to disk when overcommitted, last resort option.
Question 3: Did HA and DRS work if vCenter is down?
HA (High Availability): Yes, HA continues to function as it is configured at cluster level and works with ESXi hosts directly.
DRS (Distributed Resource Scheduler): No, DRS depends on vCenter and will not function if vCenter is down.
Question 4: What is VMware HA master election process?
When HA is enabled, one ESXi host is elected as the master.
Election process occurs when:
HA is first enabled.
Current master fails or is disconnected.
Election criteria:
Host with access to the greatest number of datastores.
If tied, the host with the highest Managed Object ID (MOID) is chosen.
No comments:
Post a Comment