[Starlingx-discuss] [ Test ] meeting minutes - 05/14/19

[Cabrales, Ada]

Agenda for 05/14
Attendees: Elio, Richo, Abraham, Cristopher, JC, JP, Maria P, Numan, abailey (Al), Fernando, Ada, Victor
 
1. Virtual environment adjustments, status - Cristopher
     We haven't been able to install a virtual env since 4 week ago.
     looks like the amount of memory to be used depends on the amount of cores present. 
     We have progress with virtual machines with 3 cores. Somebody (Brent) in the community meeting last week mentioned that was a way for modify this during deployment time. We haven't found the way of doing it. We are  still waiting for the information.
     
     Numan - not aware of a way of modifying the memory required. 
     
     https://bugs.launchpad.net/starlingx/+bug/1825814   is a change for huge pages.  
   Merged May 3. This checks if it's running on a virtual env - verify if qemu returns "true" to this check.
     
     Intel team hasn't tried with Virtual Box 
     Victor to work with Jose in order to verify the commit listed above. 
 
2. Sanity issues review, summary - Cristopher
    We are not able to deploy any bare metal configuration.  The issues are different from config to config
    Duplex: https://bugs.launchpad.net/starlingx/+bug/1828877 - Assigned to zhipeng liu - Ada to check
    Simplex:  remains on the step below - can stay there for hours
        stx-openstack       | 1.0-13-centos-stable-latest | armada-manifest               | manifest.yaml | uploading | None      |
        https://bugs.launchpad.net/starlingx/+bug/1827521  - due to ceph-rest-api - being worked right now.
    Standard: error during application-apply. Some error found about sysinv.conductor.kube_app.
    HTTP response body: {"kind":"Status","apiVersion":"v1","metadata":{},"status":"Failure","message":"secrets \"ceph-pool-kube-rbd\" not found","reason":"NotFound","details":{"name":"ceph-pool-kube-rbd","kin
      Looks like this one is the same as duplex. 

   https://bugs.launchpad.net/starlingx/+bug/1828059 - cinder volume create failed - seen by WR

    Numan's team has red status too, experiencing similar problems 

    Tomorrow in the community meeting we will begin the discussion about having all these changes bringing instability to the build. Bruce to include it in the agenda.    

3. quick round of feature testing status - All
      OVS-DPDK update change to be merged soon. - We won't be able to work with OVS as containerized service after this merge. We'll have to wait until OVS is updated.
      OVS firewall - Elio to set a meeting with Matt for clarify the usage of calico for this.
      OVS pmon - straightforward test - tested along with firewall
      JP working on adding the new networking tests to the sanity suite.
      Ceph upgrade -  26 test, 10 attempted, 2 passed, 6 blocked (steps and launchpad), 1 failed
           Thanks to Maria Y for her help.
           Abraham working on getting an extra pair of hands to help on this.
       More disk required  - have a bunch of hdd and sdd.  Ask Cristopher for them.
       Maria Y has transitioned to other project. Direct your questions to Numan, in the meantime some other person can answer.
       Containers - continue working on creation of the steps  for the tests defined.
            Numan - having good progress. Because of the ansible change, working on re-deploying labs. Changes in automation WIP.
        
        Ansible - simplex labs done. Working on changing the standard configs. In the wiki there's only instructions for running it on simplex. Ada to ask for the wiki updates on the wiki. 
               Trying the different configurations that can be made through config files. 
           JC working on updating intel's automation. 
        Openstack patch elimination - 4 tested (11 tests).
            Numan - update to be provided offline.
        QAT - not able to deploy an ISO that has the QAT code merged (because of the issues the build has been presenting). 
               Feedback for the tests proposed is welcome. 
               
4. Opens - All
      Elio - 195 tests automated so far. Slow progress due to the iso stability problems.
      Numan - try to hurry with the update for the ansible change. config_controler is not maintained anymore.
      Victor - tested the patch for the virtual env, work with Cristopher for the qemu env.