Regarding vSAN 6.7 memory consumption calculations,
I am trying to calculate the vSAN overhead / memory calculations for a 4 node VxRAIL E560F cluster. Each node has the following configuration:
CPU | VXR INTEL CPU GD 6238 2.1G 22C | 1 x Physical CPU = 22 Cores |
RAM | VXRAIL MEMORY 32GB 2933MT RDIMM | x12 32 GB RDIMM Modules |
Cache Devices for vSAN | VXR MIXED USE 800GB 2.5 CACHE | 1 x 800 GB SSD |
Capacity Devices for vSAN | VXR 1.92TB CAPACITY SATA 2.5IN SSD | 2 x 1.92 TB |
There is a KB article out: https://kb.vmware.com/s/article/2113954. In this article they describe (for example) that you can calculate an all flash configuration with 1 disc group with the following formula:
HOST_FOOTPRINT + NumDiskGroups * (DISKGROUP_FIXED_FOOTPRINT + DISKGROUP_SCALABLE_FOOTPRINT + CacheSize * CACHE_DISK_FOOTPRINT + NumCapacityDisks * CAPACITY_DISK_FOOTPRINT).
They say at the beginning that with all flash systems, the DISKGROUP_SCALABLE_FOOTPRINT = 0.5% of system memory.
In the examples they give a host with 256GB memory as an example. 0.5% of 256GB is 1,28 or 128000MB.
In the examples they have 1310 and 1228 for their DISKGROUP_SCALABLE_FOOTPRINT. Not sure how they are getting this number at all.
Maybe I have missed something fundamental here? Any assistance is very much appreciated
Thank you
Just reading it again:
1228 = hybrid
1310 = all-flash
those are different as hybrid takes the smallest of 0.5% of system memory vs 0.2% of cache disk size. In the hybrid case the cache disk is 600GB, which results in 1228.
Just reading it again:
1228 = hybrid
1310 = all-flash
those are different as hybrid takes the smallest of 0.5% of system memory vs 0.2% of cache disk size. In the hybrid case the cache disk is 600GB, which results in 1228.
Thanks for this, I was a bit perplexed by how they were getting to that number. I posted this in another group and they said:
1310 number: 1.28 GB is 1310 MB. It's in base 2, so make sure you multiply by 1024 and not 1000.
1228 number: It says that for hybrid the DISKGROUP_SCALABLE_FOOTPRINT is either 0.05% of the system memory OR 0.02% of the cache disk size, whichever is smaller. in the example, the cache disk is 600GB and the RAM is 256GB.
In that case, the two numbers are: 0.05% of 256GB = 1.28 GB, 0.02% of 600GB = 1.2 GB. So the 1.2GB is going to be used.
1.2 GB * 1024 = 1228 MB of RAM
Yeah the KB isn't very clear, I will ask the team to see if they can improve readability.
So to complete my calculation based on my scenario:
Constants:
HOST_FOOTPRINT = 7100 MB
CAPACITY_DISK_FOOTPRINT = 160 MB (ALL_FLASH)
DISKGROUP_FIXED_FOOTPRINT = 1360 MB (ALL_FLASH) *
CACHE_DISK_FOOTPRINT = 20 MB (ALL_FLASH)
DISKGROUP_SCALABLE_FOOTPRINT = 0.5% of system memory (0.5% of system memory *1024)
To calculate the vSAN memory consumption footprint, the following equations are used:
vSANFootprint = HOST_FOOTPRINT + NumDiskGroups * DiskGroupFootprint
DiskGroupFootprint = DISKGROUP_FIXED_FOOTPRINT + DISKGROUP_SCALABLE_FOOTPRINT + CacheSize * CACHE_DISK_FOOTPRINT + NumCapacityDisks * CAPACITY_DISK_FOOTPRINT
My scenario:
I am trying to calculate vSAN overhead for a cluster of 4 nodes.
Each host has x12 36GB RDIMM Modules = 432GB per node or 1.72TB for a cluster of 4 nodes.
DISKGROUP_SCALABLE_FOOTPRINT = 1,72TB = 1720GB so 0,5% of 1720 = 8.6*1024 = 8806
DiskGroupFootprint = 1360 + 8806 + 800 * 20 + 2 * 160 = 26.486
vSANFootprint = 7100 + 1 *26.486 = 7126MB
Just wanted to make 100% sure, thanks a lot!
the calculations are per host, you are doing the math per cluster. I would just use this spreadsheet, a bit easier