Bug 2151291

Summary: ODF Overview page and Overview Block & File page IOPS, Throughput & Latency values not matching due to difference in query and time intervals
Product: [Red Hat Storage] Red Hat OpenShift Data Foundation Reporter: Ramakrishnan Periyasamy <rperiyas>
Component: ocs-operatorAssignee: Mudit Agarwal <muagarwa>
Status: CLOSED NOTABUG QA Contact: Martin Bukatovic <mbukatov>
Severity: high Docs Contact:
Priority: unspecified    
Version: 4.9CC: madam, ocs-bugs, odf-bz-bot, sostapov
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2022-12-07 11:13:15 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Ramakrishnan Periyasamy 2022-12-06 15:44:43 UTC
Description of problem (please be detailed as possible and provide log
snippests):
This BZ is subset of https://bugzilla.redhat.com/show_bug.cgi?id=2026144 for tracking the enhancement on IOPS, Throughput & Latency values to match with ODF Overview page and Overview Block & File page. 

During backport verification of BZ https://bugzilla.redhat.com/show_bug.cgi?id=2124282 observed Latency difference around ~2ms, after discussion with dev came to know about the comment from Arun https://bugzilla.redhat.com/show_bug.cgi?id=2026144#c20 on why this difference or mismatch.

Copying comments from BZ 2026144#c20

ODF Overview page graphs (IOPS, Throughput and Latency) shows the sum of Read+Writes of I/O when compared with the separate Read and Write graphs in StorageSystem's -> Overview Block & File page. So both set of graphs won't have a one to one correlation.

For,
IOPS
      ODF Overview page and StorageSystem Overview page has different time intervals.

Throughput
      Underlying queries are different with different collection interval.

Latency
     Again two totally different queries (where one taking an average and the other taking sum).

For IOPS we can correct the time interval and make it same and for the other two need to check/investigate a bit further on the queries.


Version of all relevant components (if applicable):


Does this issue impact your ability to continue to work with the product
(please explain in detail what is the user impact)?
NA

Is there any workaround available to the best of your knowledge?
NA

Rate from 1 - 5 the complexity of the scenario you performed that caused this
bug (1 - very simple, 5 - very complex)?
1

Can this issue reproducible?
Yes

Can this issue reproduce from the UI?
Yes

If this is a regression, please provide more details to justify this:


Steps to Reproduce:
Install odf operator
2. Create storagesystem
3. Navigate to ODF tab under Storage, on Overview page, check the values of IOPS, Throughput, Latency from Performance card and Used Capacity value from System Capacity card on this page.
4. Match these values against the values listed in Utilization card on Block and File tab. Click on ocs-storagecluster-storagesystem hyperlink from Performace card on Overview page to navigate to Block and File tab.

Actual results:
There is difference in values for IOPS, Throughput and Latency between ODF Overview page and Overview Block & File page

Expected results:
Values of IOPS, Throughput and Latency between ODF Overview page and Overview Block & File page should match.

Additional info:
NA

Comment 2 Ramakrishnan Periyasamy 2022-12-07 11:09:37 UTC
BZ https://bugzilla.redhat.com/show_bug.cgi?id=2026144 has the fix for query and time mismatch.

Actual problem is with 4.9 UI. In latest 4.12 there is a Read and Write color coding data in StorageSystem Block and File page but in 4.9 there is no clarity on either its aggregate of both read and write or its just only write/read. 

Attached screen shot for reference, When we backport any changes in ocs-operator then corresponding UI change/update PR's too should be backported to avoid above mentioned problem.

Comment 5 Ramakrishnan Periyasamy 2022-12-07 11:13:15 UTC
Based on comment 2, 3 & 4 closing this BZ