On new install 4.0 rc5 volume sizes exceed available space

[unman]

Qubes OS version:

4.0 rc5

Affected component(s):

qubesd

---

Steps to reproduce the behavior:

Install rc5.

Expected behavior:

No qubesd warnings - possibly pools accomodate to available space.

Actual behavior:

journal filled with qubesd warning:
Sum of all thin pools exceeds size of thin pool and the size of whole volume group.
dmeventd 0 no longer monitoring thin pool qubes_dom0-pool00-tpool

General notes:

I dont know how important this is, or whether the log entries are just a distraction.
The implication seems to be that the default install allocates more than the available space - since the stated minimum requirement remains 32 GB, isn't this going to cause problems?
If I've misunderstood what's happening here, close this and accept my apologies.

---

Related issues:

[tasket]

IIUC this is a standard effect of over-provisioning and can be ignored. But you'll notice that the "no longer monitoring thin pool" messages are immediately followed with "Monitoring thin pool...".

The only thing I don't get are the messages that follow stating "Thunar: Failed to launch the volume manager". Looks like a misconfiguration of dmeventd or related tool.

What must be done to safeguard against over-allocation is covered I think by issue #1872

[tasket]

IIUC this is a standard effect of over-provisioning and can be ignored. But you'll notice that the "no longer monitoring thin pool" messages are immediately followed with "Monitoring thin pool...".

The only thing I don't get are the messages that follow stating "Thunar: Failed to launch the volume manager". Looks like a misconfiguration of dmeventd or related tool.

What must be done to safeguard against over-allocation is covered I think by issue #1872

[marmarek]

@tasket is right, over-provisioning here is expected (most of it is caused by snapshots of template's root for each VM started from it). Apparently there is no option to mute this warning, even in legitimate use cases:
https://bugzilla.redhat.com/show_bug.cgi?id=1347008
https://www.redhat.com/archives/linux-lvm/2016-April/msg00032.html

I'd like to avoid redirecting the whole stderr to /dev/null, to have logged relevant messages. Maybe some filtering grep -v ... style. It's messy.

[marmarek]

@tasket is right, over-provisioning here is expected (most of it is caused by snapshots of template's root for each VM started from it). Apparently there is no option to mute this warning, even in legitimate use cases:
https://bugzilla.redhat.com/show_bug.cgi?id=1347008
https://www.redhat.com/archives/linux-lvm/2016-April/msg00032.html

I'd like to avoid redirecting the whole stderr to /dev/null, to have logged relevant messages. Maybe some filtering grep -v ... style. It's messy.