Enthusiast
Enthusiast

How can my vSAN handle a 1 -node failure when I am using more than 50% of my combined storage?

Jump to solution

I have a 2 node vSAN on 6.5 vcenter/hosts.  Each node has a 300GB disk(cache) and 1.6TB disk(storage).  My vSAN capacity shows  as 3.3TB of storage.  That alone scares me because I that means that the data is not mirrored and either a host failure, or drive failure, would cause problems with HA.  Why is VMWare showing all of that capacity and no alarms when I am over 50% of storage used?  Is it because my disks are thin provisioned and possibly my disk usage lower than what vSAN is showing?  I am sure this is simple, but what am I missing here.  I did have a problem last year with admission control not starting up VMs after a failure when it needed to and could have, so that is turned off.  Is that the issue?

0 Kudos
1 Solution

Accepted Solutions
Leadership
Leadership

The disk space that will be used depends on the assigned policies. With the Default Policy, objects will be mirrored, i.e. 10GB of data will use 20GB (10GB on each host). Other policies will use more, or less disk of the available space. Since vSAN doesn't know which policies will be used, it shows the complete available disk space.

André

View solution in original post

0 Kudos
2 Replies
Leadership
Leadership

The disk space that will be used depends on the assigned policies. With the Default Policy, objects will be mirrored, i.e. 10GB of data will use 20GB (10GB on each host). Other policies will use more, or less disk of the available space. Since vSAN doesn't know which policies will be used, it shows the complete available disk space.

André

View solution in original post

0 Kudos
Enthusiast
Enthusiast

I am using the vSAN default storage policy to handle 1 failure.  It looks like it is just doubling everything.  So it reports that my available storage is twice what it actually is because as I start adding in more space, ir copies the data to both nodes(mirror) and would fill when it gets to 50% of what it reports.  So it reports the VMs are twice the size they actually are, etc.  I guess I am just use to dealing with actual numbers and views based on what  RAID policy is applied, etc. Thanks

0 Kudos