Bug 1894889 - Node filesystem used and total calculations are still wrong
Summary: Node filesystem used and total calculations are still wrong
Keywords:
Status: CLOSED DUPLICATE of bug 1893601
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: Management Console
Version: 4.6
Hardware: Unspecified
OS: Unspecified
medium
medium
Target Milestone: ---
: 4.7.0
Assignee: ralpert
QA Contact: Yadan Pei
URL:
Whiteboard:
: 1898609 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2020-11-05 11:18 UTC by Pablo Alonso Rodriguez
Modified: 2024-03-25 16:57 UTC (History)
4 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2020-11-19 15:07:31 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)

Description Pablo Alonso Rodriguez 2020-11-05 11:18:34 UTC
Description of problem:

This is a continuation of https://bugzilla.redhat.com/show_bug.cgi?id=1874028 , because fixes in that BZ are not enough.

Current query after BZ fix looks like this:

sum by (instance) (node_filesystem_size_bytes{fstype!~"tmpfs|squashfs",mountpoint!~"/usr|/var"} - node_filesystem_free_bytes{fstype!~"tmpfs|squashfs",mountpoint!~"/usr|/var"})

But it is not enough because:
- Mountpoint at `/sysroot` is not ignored, although it is a duplicate of `/` (this one is reported at all the reproducers I could have access to)
- Mountpoint at `/etc` is not ignored, although it is a duplicate of `/` (this one is reported at my 4.6.1 own reproducer)
- nsfs mount points are not ignored, although they should (I only had this at my 4.5 reproducers, but we need to filter them out in 4.6+ as well)

The queries should be fixed to look like:

sum by (instance) (node_filesystem_size_bytes{fstype!~"tmpfs|squashfs|nsfs",mountpoint!~"/usr|/var|/etc|/sysroot"} - node_filesystem_free_bytes{fstype!~"tmpfs|squashfs|nsfs",mountpoint!~"/usr|/var|/etc|/sysroot"})


Version-Release number of selected component (if applicable):

4.6.1 and 4.5

How reproducible:

Always in certain reproducer environments

Steps to Reproduce:
1. Open console 
2. Go to nodes --> storage
3. See that calculations are still wrong

Actual results:

Calculations wrong

Expected results:

Calculations correct

Additional info:

I can easily build both 4.6.1 and 4.5 reproducers for any data needed.

Comment 6 ralpert 2020-11-13 14:55:45 UTC
PR is complete. Monitoring still hasn't reviewed the PR, so it likely won't merge by the end of the sprint given that they're in the EU.

Comment 7 ralpert 2020-11-18 19:04:51 UTC
*** Bug 1898609 has been marked as a duplicate of this bug. ***

Comment 8 ralpert 2020-11-18 19:25:40 UTC
One of the existing queries excluded /boot|/boot/efi unlike all the others. I'm trying to find out who can confirm whether it's required there. Monitoring said they don't have domain knowledge here, and neither does the UI team.

Comment 9 ralpert 2020-11-19 15:04:54 UTC

*** This bug has been marked as a duplicate of bug 1893601 ***

Comment 10 ralpert 2020-11-19 15:07:31 UTC

*** This bug has been marked as a duplicate of bug 1893601 ***


Note You need to log in before you can comment on or make changes to this bug.