[Starlingx-discuss] Agenda: Weekly StarlingX non-OpenStack Distro meeting, 4/24
Xie, Cindy
cindy.xie at intel.com
Wed Apr 24 13:39:02 UTC 2019
Agenda & Notes for 4/24 meeting:
- Ceph upgrade status
1. Ceph dev build validation status (Fernando)
Pre-merged basic test ran.
__________________________________________________________________________________________________
# | Test ID | Comments
1 | STOR_TIER_005 | Per Frank Miller: ...deferred until task 30351 under the HELM SB 2003909 is completed
2 | STOR_TIER_006 | Per Frank Miller: ...deferred until task 30351 under the HELM SB 2003909 is completed
3 | STOR_TIER_007 | Per Frank Miller: ...deferred until task 30351 under the HELM SB 2003909 is completed
4 | STOR_TIER_008 | Per Frank Miller: ...deferred until task 30351 under the HELM SB 2003909 is completed
**SB- https://storyboard.openstack.org/#!/story/2003909
7 | STOR_PROCESS_011 | - WIP - Emails have been back and forward and test case has been re-worked.
| Pending questions:
| @Daniel.
- Does StarlingX Ceph solution require at least ceph-mgr, or 2 (active and standby) at the same time? It seems that only one ceph-mgr also works.
[Daniel]: ceph-mgr has the HA mechanism built-in so it's not depends on HA. Needs restful plug-in to restart the service. race condition for port 5001. Working on fixing the issues more testing required. root-cause identified: the services itself is not restarted only the plug-in is up. If we have multi-layer implemented, it can restart the service. Impact: minimal due to the fact if we have 2 serivies running and 1 of them is not responsing, then it still working. Ceph-mgr REST API has port 5001, race condition of 2 ceph-mgr to access the same ports.
AR: Daniel to fix the issue before we call the patch can be merged. ETA: patch to be ready by tomorrow.
- In the case I described above, once ceph-mgr on controller-1 took over the active role, ceph-mgr on controller-0 hardly won the chance to be "active" any more. Is it reasonable?
- By "systemctl status sm.service | grep ceph-mgr", we know ceph-mgr daemon is managed by sm.service, but why sm.service did not restart ceph-mgr WHEN there is still a ceph-mgr on another controller?
| @Daniel/Yong.
- For STEP 8. Need guidance about what ceph folder/filename must be renamed to get the ceph monitor process killed and never come back. It seems that renaming the ceph services from path "/usr/lib/systemd/system" is not working for this quest.
Fernando needs to rename the binary (path?) Daniel to provide guiance using email for specific file names.
- Do you believe is this still a valid test case? We can focused on kill the service but Is this going to happen to a customer, Is this a real scenario?
8 | STOR_PROCESS_012 | - WIP - Emails have been back and forward and test case has been re-worked.
| Pending questions:
| This test case was executed almost completely, the only step missing is the "Step 9" when killing the osd process and never back. I need @Daniel/Yong/Tingkie feedback in order to kill ceph service to never come back once they are down.
[Fernando] same quesiton as of test case 011, need to rename the correct binaries.
17 | STOR_FAULT_023 | Not Attempted
20 | STOR_PART_030 | Not Attempted
AR: Fernando to run the auto-sanity for the same binary after the cengen built. ETA: end of today. Fernando please send the Sanity detail failures to the list so that we can help to do the failure analysis.
2. test plan clarification (Fernando/Daniel) - covered in (1)
3. Patch rebase (Daniel)
https://review.openstack.org/#/q/topic:ceph-mimic-upgrade+(status:open+OR+status:merged)
Daniel rebased all patches yeserday and working on the patches for ceph-mgr race condition issue.
- QAT driver upgrade status
1. Dev status (base rebase, QAT driver load, ISO generation, etc) (Haitao)
Still working on setting up the environment in offshore ODC and meet deployment issue in the public network. Helping ODC to deploy the ISO from SH lab. QAT devices can be shown from Horizon and we can start the VM esting to access the QAT driver from tenant.
2. validation prepration (Ricardo)
Ricardo is waiting for ISO generated from ODC. test cases discussion between Ricardo and Haitao for QAT tests.
- libvirt/qemu patch reduction PR merge status (Jim/Dean)
2 PR for Libvirt and 1 for Qemu posted. Dean to review the PR.
- Opens (all)
-----Original Message-----
From: Xie, Cindy
Sent: Tuesday, April 23, 2019 9:06 PM
To: starlingx-discuss at lists.starlingx.io
Cc: 'Rowsell, Brent' <Brent.Rowsell at windriver.com>; Wold, Saul <saul.wold at intel.com>; Hernandez Gonzalez, Fernando <fernando.hernandez.gonzalez at intel.com>; Badea, Daniel <Daniel.Badea at windriver.com>; Wang, Hai Tao <hai.tao.wang at intel.com>; Perez, Ricardo O <ricardo.o.perez at intel.com>; Somerville, Jim <Jim.Somerville at windriver.com>; 'Khalil, Ghada' <Ghada.Khalil at windriver.com>; Troyer, Dean <dean.troyer at intel.com>
Subject: Agenda: Weekly StarlingX non-OpenStack Distro meeting, 4/24
Agenda for 4/24 meeting:
- Ceph upgrade status
1. Ceph dev build validation status (Fernando)
2. test plan clarification (Fernando/Daniel)
3. Patch rebase (Daniel)
- QAT driver upgrade status
1. Dev status (base rebase, QAT driver load, ISO generation, etc) (Haitao)
2. validation prepration (Ricardo)
- libvirt/qemu patch reduction PR merge status (Jim/Dean)
- Opens (all)
-----Original Appointment-----
From: Xie, Cindy
Sent: Monday, November 5, 2018 2:27 PM
To: Xie, Cindy; Waheed, Numan; Hu, Yong; starlingx-discuss at lists.starlingx.io; Liu, ZhipengS; Shang, Dehao; Wold, Saul; Lin, Shuicheng; Zhu, Vivian; Somerville, Jim; Sun, Austin; 'Khalil, Ghada'; Jones, Bruce E; Troyer, Dean; 'Rowsell, Brent'
Cc: 'Poncea, Ovidiu'; Gomez, Juan P; Lara, Cesar; Fang, Liang A; Cobbley, David A; 'Chen, Jacky'; Perez Rodriguez, Humberto I; Martinez Monroy, Elio; 'Seiler, Glenn'; Arce Moreno, Abraham; 'Waines, Greg'; 'Eslimi, Dariush'; 'Hellmann, Gil'; Shuquan Huang; 'Young, Ken'; Perez Carranza, Jose; Hu, Wei W; Martinez Landa, Hayde; Armstrong, Robert H
Subject: Weekly StarlingX non-OpenStack Distro meeting
When: Wednesday, April 24, 2019 6:00 AM-7:00 AM (UTC-08:00) Pacific Time (US & Canada).
Where: https://zoom.us/j/342730236
. Cadence and time slot:
o Wednesday 9AM Winter EDT (10PM China time, US PDT Winter time 6AM)
. Call Details:
o Zoom link: https://zoom.us/j/342730236
o Dialing in from phone:
o Dial(for higher quality, dial a number based on your current location): US: +1 669 900 6833 or +1 646 876 9923
o Meeting ID: 342 730 236
o International numbers available: https://zoom.us/u/ed95sU7aQ
. Meeting Agenda and Minutes:
o https://etherpad.openstack.org/p/stx-distro-other
More information about the Starlingx-discuss
mailing list