[Starlingx-discuss] CentOS 7.6 rebase feature testing

Mon Feb 4 14:59:45 UTC 2019

Thanks Shuicheng. Is there a need to actually upversion the qat drivers? Are there new features or bug fixes that are of interest to StarlingX?

I suggest you review the release notes and make a decision with the distro.other team leads before proceeding with this item.

Ghada

From: Lin, Shuicheng [mailto:shuicheng.lin at intel.com]
Sent: Wednesday, January 30, 2019 7:34 PM
To: Khalil, Ghada; Waheed, Numan; Perez, Ricardo O; Xie, Cindy; Cabrales, Ada
Cc: starlingx-discuss at lists.starlingx.io
Subject: RE: CentOS 7.6 rebase feature testing

Hi all,
Story is created to track this task:
https://storyboard.openstack.org/#!/story/2004901

Best Regards
Shuicheng

From: Lin, Shuicheng [mailto:shuicheng.lin at intel.com]
Sent: Thursday, January 31, 2019 7:56 AM
To: Khalil, Ghada <Ghada.Khalil at windriver.com<mailto:Ghada.Khalil at windriver.com>>; Waheed, Numan <Numan.Waheed at windriver.com<mailto:Numan.Waheed at windriver.com>>; Perez, Ricardo O <ricardo.o.perez at intel.com<mailto:ricardo.o.perez at intel.com>>; Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>
Cc: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: Re: [Starlingx-discuss] CentOS 7.6 rebase feature testing

Hi Ghada,
QAT driver is not upgraded yet, since there is no build failure with CentOS 7.6.
I will have a check with it after Chinese New Year's holiday, and plan to do the upgrade in master.

Best Regards
Shuicheng

From: Khalil, Ghada [mailto:Ghada.Khalil at windriver.com]
Sent: Wednesday, January 30, 2019 11:01 PM
To: Waheed, Numan <Numan.Waheed at windriver.com<mailto:Numan.Waheed at windriver.com>>; Perez, Ricardo O <ricardo.o.perez at intel.com<mailto:ricardo.o.perez at intel.com>>; Lin, Shuicheng <shuicheng.lin at intel.com<mailto:shuicheng.lin at intel.com>>; Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>
Cc: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: RE: CentOS 7.6 rebase feature testing

Shuicheng,
Can you confirm if the qat17 driver was updated?  I didn't see it on the task list.

To the test team,
I suggest you look at the summary of changes highlighted by Shuicheng below as well as the tasks in the stories (as they show the packages that have changed) and choose the test-cases and test configurations accordingly.

Based on the changes listed in https://storyboard.openstack.org/#!/story/2004521

-        Sanity/Basic Regression on standard and low-latency profile

o   Install/lock/unlock all node types (controller/compute/storage)

o   VM operations

-        Niantic NIC tested as a mgmt/infra and data interface

-        Fortville NIC tested as a mgmt/infra and data interface

-        Mellanox NIC tested as a mgmt/infra interface, but not a data interface (as mlx driver DPDK support is disabled)

o   Include cable pulls for all NIC types

-        Hardware TPM

-        IMA/Integrity

-        DRBD sync - should be covered by duplex/multi-node test-cases (lock/unlock/reinstall controller-1)

Based on the changes listed in https://storyboard.openstack.org/#!/story/2004521

-        https config (changes in haxproxy)

-        snmp traps (changes in net-snmp)

-        ldap user password changes (changes in openldap)

-        efi pxe install (changes to grub2)

-        The remaining packages are covered by the normal install and basic system operations

I think the above would make a good regression suite for an OS upgrade in the future as well. I encourage others on the mailing list to review/comment as well.

Regards,
Ghada

PS: The TCs below are heavily geared towards neutron. They don't provide sufficient coverage given the areas of churn.

From: Waheed, Numan
Sent: Wednesday, January 30, 2019 9:08 AM
To: Perez, Ricardo O; Lin, Shuicheng; Khalil, Ghada; Xie, Cindy; Cabrales, Ada
Subject: RE: CentOS 7.6 rebase feature testing

Hi Ricardo,

Thanks for providing the list of test cases. These are mostly networking related test cases. Can you please let me know which NICs you are using in your labs?

I would also suggest to add some tests that are specific to OS e.g. password related tests, Backup and Restore etc.
Are you also going to do any low-latency setup?
I would also suggest to add some Nova related tests.

Thanks,

Numan.

From: Perez, Ricardo O <ricardo.o.perez at intel.com<mailto:ricardo.o.perez at intel.com>>
Sent: January-29-19 6:08 PM
To: Waheed, Numan <Numan.Waheed at windriver.com<mailto:Numan.Waheed at windriver.com>>; Lin, Shuicheng <shuicheng.lin at intel.com<mailto:shuicheng.lin at intel.com>>; Khalil, Ghada <Ghada.Khalil at windriver.com<mailto:Ghada.Khalil at windriver.com>>; Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>
Subject: RE: CentOS 7.6 rebase feature testing

Hi Numan,

These are the tests that we are planning to run on CentOS 7.6 Testing:

Live-migration: Instance is Scheduling on socket where vswitch is running (on destination host)
Lock & Unlock the compute
Off-line static configuration for "External OAM" Interface
On-line Static configuration validation: swact controllers should be rejected until controller-0 config for "External OAM" Inter
Provider network Down Alarm
Reject changing interface MTU size to values smaller than MTU of provider network
Verify administrator is able to set alter and query System name Via CLI and GUI
Verify alarm generation for neutron L3 agent scheduling states
Verify alarm generation for neutron provider network state
Verify appropriate values should be used for modifying of interfaces (CLI, GUI)
Verify ethernet MGMT interface is updated successfully on controller.
Verify ethernet OAM interface is updated successfully on controller.
Verify that all providernettypes can be shown via Neutron REST API
Verify that associating provider network with an interface is rejected if MTU of the interface is smaller than MTU of provider n
Verify that changing the physical interface of MGMT network is permitted
Verify that changing the physical interface of OAM network is permitted
Verify that hosts can be shown via Neutron REST API
Verify that interfaces parameters can be modified via API reques
Verify that interfaces ports data can be modified via API.
Verify that providernet can be created/updated/deleted via Neutron REST API
Verify that providernetranges can be created/updated/deleted via Neutron REST API
Verify that subnet can be created and subnet attributes can be modified via Neutron REST API
Verify that System name can be modified via CLI and GUI
Verify that the MTU size is displayed correctly for the infra interface after it is changed.
Verify that unlocked powered off host can not be deleted (CLI, GUI)

The configurations that we are pretending to use are:
P1 - 2+2+2 (External Storage)
P2 - Duplex

If someone has any other test that we have to consider or add to this list, just let us know.

Regards
-Ricardo

From: Waheed, Numan [mailto:Numan.Waheed at windriver.com]
Sent: Monday, January 28, 2019 7:49 AM
To: Lin, Shuicheng <shuicheng.lin at intel.com<mailto:shuicheng.lin at intel.com>>; Khalil, Ghada <Ghada.Khalil at windriver.com<mailto:Ghada.Khalil at windriver.com>>; Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>
Cc: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: RE: CentOS 7.6 rebase feature testing

Hi Ada,

Which tests you are planning to run for CentOS 7.6 testing? Which configs you are using?

Thanks,

Numan.

From: Lin, Shuicheng <shuicheng.lin at intel.com<mailto:shuicheng.lin at intel.com>>
Sent: January-28-19 1:28 AM
To: Khalil, Ghada <Ghada.Khalil at windriver.com<mailto:Ghada.Khalil at windriver.com>>; Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>
Cc: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: Re: [Starlingx-discuss] CentOS 7.6 rebase feature testing

Hi Ghada/Numan,
1 more patch is added for feature branch.
https://review.openstack.org/633431

Best Regards
Shuicheng

From: Lin, Shuicheng [mailto:shuicheng.lin at intel.com]
Sent: Monday, January 28, 2019 10:20 AM
To: Khalil, Ghada <Ghada.Khalil at windriver.com<mailto:Ghada.Khalil at windriver.com>>; Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>
Cc: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: Re: [Starlingx-discuss] CentOS 7.6 rebase feature testing

Hi Ghada/Numan,
All code for CentOS 7.6 upgrade has been merged to centos76 feature branch.
Please have a try with it, and notify me if there is any problem.
Manifest file I used is attached, it is with last week's master code + feature branch code for stx-integ/stx-upstream/stx-tools/stx-root.

Here is a simple summary for CentOS 7.6 upgrade:
30 srpm & 650+ rpm upgraded.
Kernel version upgraded to 3.10.0-957.1.3 for both std and rt kernel.
Out of tree driver upgraded/updated: i40e/i40evf/ixgbe/ixgbevf/tpm/integrity/mellanox/rdma-core/libibverbs.
Mellanox driver support in DPDK is disabled. (DPDK/OpenVswitch need be upgraded later to support Mellanox driver again.)
You could get more info from below stories:
Kenrel upgrade story: https://storyboard.openstack.org/#!/story/2004521
Srpm & rpm upgrade story: https://storyboard.openstack.org/#!/story/2004522
Mellanox driver support: https://storyboard.openstack.org/#!/story/2004743

Best Regards
Shuicheng

From: Khalil, Ghada [mailto:Ghada.Khalil at windriver.com]
Sent: Thursday, January 24, 2019 3:15 AM
To: Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>; Lin, Shuicheng <shuicheng.lin at intel.com<mailto:shuicheng.lin at intel.com>>
Cc: Waheed, Numan <Numan.Waheed at windriver.com<mailto:Numan.Waheed at windriver.com>>; Jones, Bruce E <bruce.e.jones at intel.com<mailto:bruce.e.jones at intel.com>>; starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: RE: CentOS 7.6 rebase feature testing

Sorry for the confusion. Yes Numan will wait until all the pending patches are merged in the feature branch. Is this planned for the end of next week (before Chinese New Year holidays)?

From: Xie, Cindy [mailto:cindy.xie at intel.com]
Sent: Wednesday, January 23, 2019 2:10 PM
To: Khalil, Ghada; Cabrales, Ada; Lin, Shuicheng
Cc: Waheed, Numan; Jones, Bruce E; starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: RE: CentOS 7.6 rebase feature testing

Ghada,
In today's meeting, Ken mentioned that he'd like to see all pending patches merged to f/CentOS7.6 for Numan to start. But if Numan wants to start earlier, that's fine. @Shuicheng should be able to send out the build instructions.

Thx. - cindy

From: Khalil, Ghada [mailto:Ghada.Khalil at windriver.com]
Sent: Wednesday, January 23, 2019 11:05 AM
To: Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; Cabrales, Ada <ada.cabrales at intel.com<mailto:ada.cabrales at intel.com>>
Cc: Waheed, Numan <Numan.Waheed at windriver.com<mailto:Numan.Waheed at windriver.com>>; Jones, Bruce E <bruce.e.jones at intel.com<mailto:bruce.e.jones at intel.com>>; starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: CentOS 7.6 rebase feature testing

Hi Cindy,
My understanding from the F2F meeting last week that you already discussed with Ada testing the CentOS 7.6 rebase image (feature branch) starting Feb 1 while your team is off for Chinese New Year's.

Can you please send the instructions for building the feature branch to the stx mailing list? Numan will try to have an image built to try it out in one of the WR labs (time-permitting).
I expect the feature branch is rebased to the latest from master.

Hi Ada,
Given that there are driver upgrades for i40e & ixgbe, I assume you are planning to test on baremetal hardware, not just in the virtual env. Are your baremetal labs setup now? What NIC types do you have covered in your labs?

Thanks,
Ghada

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.starlingx.io/pipermail/starlingx-discuss/attachments/20190204/2f734723/attachment-0001.html>