09:04:56 <chigang_> #startmeeting Compass4nfv weekly meeting
09:04:56 <collabot> Meeting started Wed Nov 15 09:04:56 2017 UTC.  The chair is chigang_. Information about MeetBot at http://wiki.debian.org/MeetBot.
09:04:56 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
09:04:56 <collabot> The meeting name has been set to 'compass4nfv_weekly_meeting'
09:05:44 <chigang_> #topic roll call
09:05:50 <chigang_> #info Justin
09:05:52 <yifei> #info Yifei
09:05:59 <Harry_Huang> #info harry
09:06:42 <hw_wutianwei> #info tianwei wu
09:07:08 <chigang_> #topic update from Release meeting
09:07:49 <chigang_> #info Fraser MS3.0 and Euphrates 5.1 release on Dec 15
09:08:31 <chigang_> kvm and ovs_dpdk scenarios are still no t stable in Euphrates
09:09:14 <Harry_Huang> kvm still have restart issue
09:09:42 <Harry_Huang> and dpdk has same restart code
09:10:02 <chigang_> Harry_Huang: yes, kvm failed log in #link https://build.opnfv.org/ci/job/compass-deploy-baremetal-daily-euphrates/256/consoleFull
09:10:46 <yifei> can we add e.g. wait 3 minutes?
09:11:57 <chigang_> #info kvm and ovs_dpdk scenarios failed because ansible lost compute nodes link after nodes reset
09:12:27 <chigang_> any other suggest?
09:12:41 <chigang_> any other suggestion?
09:13:45 <chigang_> yifei: how long does it wait?
09:14:10 <Harry_Huang> it's a async job
09:14:20 <yifei> chigang_: it waits for the system up
09:14:35 <yifei> chigang_: not some time now
09:15:03 <chigang_> https://www.irccloud.com/pastebin/dMLqZiml/reboot_node_code_section
09:15:18 <chigang_> I check the code, it waits 10 minutes
09:16:21 <chigang_> it works fine in HUAWEI-POD2, but fails in HUAWEI-POD1 sometimes
09:16:23 <Harry_Huang> here restart code contains two parts
09:16:54 <Harry_Huang> first is simple restart second is waiting
09:18:00 <chigang_> Harry_Huang: go on
09:19:38 <Harry_Huang> first part is a "fire and go" which is async time period of 1 sec
09:20:23 <Harry_Huang> ansible still connect to this machine in this 1 s
09:21:37 <chigang_> because there are different blade type in HUAWEI-POD1 and POD2, maybe it depends on blade's reset time.
09:22:21 <hw_wutianwei> maybe, pod1 are 2288 and pod2 are 1288
09:23:13 <chigang_> the performance of POD2 is better than POD1
09:23:24 <chigang_> hw_wutianwei: thanks
09:23:32 <yifei> does dpdk scenario have the same situation that successes on pod2, fails on pod1?
09:24:01 <Harry_Huang> one more thing to notice: there is a "sleep 2" in this 1 s async task
09:25:39 <hw_wutianwei> KVM often fail on pod2
09:26:26 <chigang_> I suppose we can try to set timeout=1200s delay=10s, it will reduce the networks traffic.
09:28:22 <chigang_> We can first try it to see if it is due to the performance of the restart time.
09:29:38 <chigang_> how about it?
09:30:19 <Harry_Huang> we can test it just on pod2
09:30:57 <yifei> ok
09:31:05 <chigang_> sure, just checkpick it into Euphrates branch.
09:31:40 <chigang_> #topic Tasks assignment in Fraser release
09:32:15 <chigang_> #undo
09:32:15 <collabot> Removing item from minutes: <MeetBot.ircmeeting.items.Topic object at 0x2e6b490>
09:33:13 <chigang_> #info set timeout=1200s delay=10s in Euphrates kvm scenarios to check if it is due to the performance of the restart time.
09:33:40 <chigang_> #topic Tasks assignment in Fraser release
09:34:17 <chigang_> #info Fraser tasks are create in #link https://wiki.opnfv.org/display/compass4nfv/Compass4nfv+Fraser+release+plan
09:35:22 <chigang_> #info Please add your name  in Own field if you are doing it.
09:35:46 <yifei> OK
09:36:21 <chigang_> #info If the priority is not appropriate, you can adjust yourself
09:37:28 <chigang_> the wiki page is created according to https://etherpad.opnfv.org/p/compass_release_fraser
09:38:00 <chigang_> you can get more information from #link https://etherpad.opnfv.org/p/compass_release_fraser
09:39:07 <chigang_> #info If the JIRA is missed, please raise a JIRA and paste the JIRA into the WIKI page #link https://wiki.opnfv.org/display/compass4nfv/Compass4nfv+Fraser+release+plan
09:40:02 <chigang_> if I missed some tasks, please add them :)
09:41:06 <chigang_> yifei: I saw your updates, Great jobs
09:41:49 <yifei> chigang_: THX. Some of them have been done
09:42:05 <chigang_> yifei: super hero in Compass
09:43:08 <chigang_> if there are unclear for some tasks, please mail me.
09:43:28 <chigang_> any comments for this topic?
09:44:20 <chigang_> #topic POD resources update for Fraser release
09:45:33 <chigang_> #info there are 3 physical POD for Compass Fraser release development
09:46:54 <chigang_> #info Intel POD17, Jack Morgen assigns it for Compass #link https://wiki.opnfv.org/display/pharos/Intel+POD17
09:47:49 <chigang_> Intel POD17 VPN access will sent to Yifei and me.
09:48:56 <chigang_> yifei: please follow the status of this POD, it isn't online now.
09:49:14 <yifei> chigang_: OK
09:50:22 <chigang_> #info the 2nd POD is HUAWEI-POD1 in Huawei Hosting
09:51:23 <chigang_> #info the 3rd POD is HUAWEI ARM POD,  it still in progress, maybe online at the end of Dec.
09:52:16 <chigang_> Taishan 2280/256G memory/4T disk/10GE,  it is well performance
09:52:36 <chigang_> hw_wutianwei: if some wrong, please correct me.
09:53:09 <hw_wutianwei> chigang_: you are right
09:53:28 <chigang_> hw_wutianwei: thanks.
09:54:19 <chigang_> that's all topics from my side, any others?
09:54:28 <yifei> no
09:55:01 <Harry_Huang> no
09:55:29 <chigang_> fine, it is dinner time. bye.
09:55:41 <chigang_> #endmeeting