Allocable Storage Value Incorrect (not matching used, available resources)

[Nathan Koterba]

Perhaps I have something mis-configured, but when I view my compute services (as admin) under:

System Panel->Services

My compute nodes are showing an erroneous value for allocable storage.

Here is my output:
nova-compute
( cloudnc3 )
Hypervisor: QEMU( rdtscp, x2apic, xtpr, tm2, est, vmx, ds_cpl, monitor, pbe, tm, ht, ss, acpi, ds, vme)
Allocable Cores: 16 (0 Used, 4 Physical/Virtual)
Allocable Storage: 9.8TB (13.0GB Used, 144.0GB Physical)
System Ram: 7GB (242MB Used)
Enabled True

I don't know where it's getting the 9.8 TB value but my physical machine only has 144 GB (correctly shown) and I am using 13 GB (again correct).

Also, my allocable cores is 16, whereas my physical machine only has 4 (non Hyper-threading capable). I'm not sure if this allows multiple instances to share a single core. Where is the value of 16 coming from? And is there any way to set it to equal the actual physical cores?

[Jay Pipes (jaypipes)]

I get *exactly* the same output on my local devstack: 9.8 TB. However, I only have 1.5 TB of storage on this machine. Also, I see the same incorrect calculation of "allocable" cores.

Here is what syspanel says:

    Hypervisor: QEMU( rdtscp, pdpe1gb, xtpr, tm2, est, vmx, ds_cpl, monitor, pbe, tm, ht, ss, acpi, ds, vme)
    Allocable Cores: 16 (0 Used, 12 Physical/Virtual)
    Allocable Storage: 9.8TB (12.0GB Used, 1.5TB Physical)
    System Ram: 23GB (1GB Used)

My machine actually has the following:

12 cores, 24GB RAM and 1.5 TB of storage.

So, something seems off. :)

[Gabriel Hurley (gabriel-hurley)]

This is fixed insofar as the broken code in openstackx that generated that value is now no longer used. Nova's API, however, does not return this data, so all in all it's a wash until we can get the data back from Nova in the future. The bug is gone, though.