Bug 2074019

Summary: [ovn-controller] nb_cfg value over a certain value causes erroneous update to SB DB
Product: Red Hat Enterprise Linux Fast Datapath Reporter: Martin Kennelly <mkennell>
Component: ovn22.03Assignee: Ales Musil <amusil>
Status: VERIFIED --- QA Contact: Ehsan Elahi <eelahi>
Severity: medium Docs Contact:
Priority: medium    
Version: FDP 22.LCC: amusil, ctrautma, dceara, jiji, jishi, mmichels
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: All   
Whiteboard:
Fixed In Version: ovn22.03-22.03.0-220.el8fdp Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 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 Martin Kennelly 2022-04-11 11:51:04 UTC 
Description of problem:
I have noticed a value over 32 bits of information to nb_cfg field in NB DB NB_Global table causes unexpected behaviour from ovn-controller.

Version-Release number of selected component (if applicable):
ovn-nbctl 21.06.1
Open vSwitch Library 2.15.90
DB Schema 5.32.0

How reproducible:
Always


Steps to Reproduce:
1. Deploy ovsdb-monitor and start to monitor OVN_Southbound or OVN_Northbound.
2. ovn-nbctl set NB_Global . nb_cfg=4294967297
3. Each chassis_private's nb_cfg field gets set to 1 and shortly after to the correct value 4294967297. Hv_cfg field gets set to 1 and shortly after to the correct value 4294967297.

Actual results:
Two updates are sent to SB DB to the ovn-controllers respective chassis_private table row.

Expected results:
I should be able to add information up to 64 bits signed to nb_cfg.
Ovn-controllers should only update SB DBs table chassis_private nb_cfg field once per update of nb_cfg field in NB DBs NB_Global table.

This is impacting a PR for me as I depend on knowing what value was previously set.

Additional info:
Comment 1 Dumitru Ceara 2022-04-11 13:01:31 UTC 
I didn't look too much into this but I suspect it's related to the mismatch between uint64_t and int64_t in ovn-controller, e.g.:

In ofctrl-seqno:

https://github.com/ovn-org/ovn/blob/b22684d4e37c3dd205a959925b6e3f1827bbe045/controller/ofctrl-seqno.h#L28

In ovn-controller.c:
https://github.com/ovn-org/ovn/blob/b22684d4e37c3dd205a959925b6e3f1827bbe045/controller/ovn-controller.c#L870
https://github.com/ovn-org/ovn/blob/b22684d4e37c3dd205a959925b6e3f1827bbe045/controller/ovn-controller.c#L887
Comment 3 Ales Musil 2023-07-13 09:17:50 UTC 
Patch posted: https://patchwork.ozlabs.org/project/ovn/patch/20230713082335.386633-1-amusil@redhat.com/
Comment 4 OVN Bot 2023-07-18 04:08:34 UTC 
ovn23.06 fast-datapath-rhel-8 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223475
ovn23.06 fast-datapath-rhel-9 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223476
ovn23.03 fast-datapath-rhel-8 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223478
ovn23.03 fast-datapath-rhel-9 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223479
ovn22.12 fast-datapath-rhel-8 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223480
ovn22.12 fast-datapath-rhel-9 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223481
ovn22.09 fast-datapath-rhel-8 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223482
ovn22.09 fast-datapath-rhel-9 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223483
ovn22.06 fast-datapath-rhel-8 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223484
ovn22.06 fast-datapath-rhel-9 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223485
ovn22.03 fast-datapath-rhel-9 clone created at https://bugzilla.redhat.com/show_bug.cgi?id=2223486
Comment 7 Ehsan Elahi 2023-07-31 13:56:12 UTC 
Reproduced on: 

ovn22.03-22.03.0-219.el8fdp.x86_64
ovn22.03-host-22.03.0-219.el8fdp.x86_64
ovn22.03-central-22.03.0-219.el8fdp.x86_64
openvswitch2.17-2.17.0-108.el8fdp.x86_64
openvswitch-selinux-extra-policy-1.0-31.el8fdp.noarch

[root@....] # ovn-nbctl set NB_Global . nb_cfg=4294967297
[root@....] # ovsdb-tool show-log -mmm /var/lib/ovn/ovnnb_db.db
---------
database records trimmed 
---------
record 7: 2023-07-31 13:40:18.205 "ovn-northd"
  table NB_Global row 651e010c (651e010c) diff:
    hv_cfg=1
    hv_cfg_timestamp=1690810818204

record 8: 2023-07-31 13:40:18.205 "ovn-northd"
  table NB_Global row 651e010c (651e010c) diff:
    hv_cfg=4294967297
    hv_cfg_timestamp=1690810818205

<====== 2 separate records are added as mentioned in the BZ description. 


Verified on:

ovn22.03-22.03.0-220.el8fdp.x86_64
ovn22.03-host-22.03.0-220.el8fdp.x86_64
ovn22.03-central-22.03.0-220.el8fdp.x86_64
openvswitch2.17-2.17.0-108.el8fdp.x86_64
openvswitch-selinux-extra-policy-1.0-31.el8fdp.noarch


[root@....] # ovn-nbctl set NB_Global . nb_cfg=4294967297
[root@....] # ovsdb-tool show-log -mmm /var/lib/ovn/ovnnb_db.db
---------
database records trimmed 
---------
record 7: 2023-07-31 13:48:54.220 "ovn-northd"
  table NB_Global row 66d21016 (66d21016) diff:
    hv_cfg=4294967297
    hv_cfg_timestamp=1690811334220


<========= Only single record added