Bug 2126397 - many KubeVirtComponentExceedsRequestedMemory alerts in Firing state
Summary: many KubeVirtComponentExceedsRequestedMemory alerts in Firing state
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Container Native Virtualization (CNV)
Classification: Red Hat
Component: Virtualization
Version: 4.11.0
Hardware: Unspecified
OS: Unspecified
urgent
high
Target Milestone: ---
: 4.12.0
Assignee: Jed Lejosne
QA Contact: Akriti Gupta
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2022-09-13 11:07 UTC by Kedar Bidarkar
Modified: 2023-01-24 13:40 UTC (History)
3 users (show)

Fixed In Version: hco-bundle-registry-container-v4.12.0-714
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2023-01-24 13:40:50 UTC
Target Upstream Version:
Embargoed:

Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Github kubevirt kubevirt pull 8613 0 None Merged Bump memory requests according to values seen in testing 2022-11-02 09:41:37 UTC
Github kubevirt kubevirt pull 8709 0 None Merged [release-0.58] Bump memory requests according to values seen in testing 2022-11-15 14:09:34 UTC
Red Hat Issue Tracker CNV-21231 0 None None None 2022-10-27 11:50:51 UTC
Red Hat Product Errata RHSA-2023:0408 0 None None None 2023-01-24 13:40:57 UTC

Description Kedar Bidarkar 2022-09-13 11:07:53 UTC 
This bug was initially created as a copy of Bug #2078545

I am copying this bug because: 



Description of problem:
 There are many alerts KubeVirtComponentExceedsRequestedMemory in Firing state on a cluster, some of them appeared right after installation.
 May be memory should be adjusted for kubevirt system pods?

Version-Release number of selected component (if applicable):
 4.11


Actual results:
 Alerts are constantly firing, some of them triggered right after cluster installation. Screenshot attached

Expected results:
 kubevirt system pods requested enough memory and alert is not triggered
Comment 1 Denys Shchedrivyi 2022-10-05 21:06:52 UTC 
 I verified it on CNV v4.12.0-548 - the alert still present for some of components (virt-api and virt-operator)

 As I understand, in the fix the requested memory was increased to 250Mb, but it is not enough for all componnets, here for example the current maximums from my cluster:

virt-api = 260Mb
virt-handler = 130Mb
virt-controller = 140Mb

Regarding virt-operator - I saw the memory usage upto 450Mb (with peaks ~570Mb).
The initial value was around 200Mb, but it increased on 15-20Mb every time when I updated Kubevirt through HCO. 
Opened another BZ: https://bugzilla.redhat.com/show_bug.cgi?id=2132473
Comment 2 Denys Shchedrivyi 2022-10-17 17:55:56 UTC 
Moving this BZ to Assigned state because virt components memory requests should be adjusted:

PSI cluster:
virt-api        - 300Mb
virt-controller - 140Mb
virt-handler    - 210Mb 
virt-operator   - 380Mb (There is Bug 2132473 with virt-operator which increases memory usage on every Kubevirt CR update, so it is difficult to predict how much memory it consumes by default)

BM cluster:
virt-api        - 285Mb
virt-controller - 200Mb
virt-handler    - 190Mb
virt-operator   - 290Mb
Comment 10 errata-xmlrpc 2023-01-24 13:40:50 UTC 
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Important: OpenShift Virtualization 4.12.0 Images security update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2023:0408
Note You need to log in before you can comment on or make changes to this bug.