Hi,
Below are some meeting minutes for the meeting yesterday.
Attendees:
- openEuler: Fengguang, Kaitian
- Huawei: Nan, Jonathan, Jerome
- Linaro: Milosz, Anmar, Ryan, Yongqin, Jammy
Notes:
- Fengguang did some introduction about CompassCI testing framework
- Fengguang summarized the major benefits of LAVA/LKFT to CompassCI
- LAVA dispatcher can support many device types
- LKFT test definitions can be wrapped in lkp_tests (both lkp jobs
and lava jobs are in YAML format)
- Anmar asked if we can wrap lkp tests in LKFT, which is also possible
according to Fengguang. But Fengguang's recommendation is to wrap LKFT in
lkp, because LAVA has some limitations to support the data model of
CompassCI.
- Anmar asked the database support in CompassCI, and Fengguang shared
that ElasticSearch is used now as the database backend to store data
- Yongqin asked if there is any plan to support Android by CompassCI.
There is no active plan at this moment, and since CompassCI is part of
openEuler community, the server OS is the focus now.
- The reports for Android Common Kernel testing can be found at [1]
- The lkq-dev mailing list [2] can be used for further discussions.
[1] https://qa-reports.linaro.org/android-lkft/
[2] https://op-lists.linaro.org/mailman/listinfo/lkq-dev
Regards,
Jammy
On Sun, 15 Nov 2020 at 13:19, Jammy Zhou via Lkq-dev <
lkq-dev(a)op-lists.linaro.org> wrote:
> Hi All,
>
> CompassCI [1] is a new CI testing framework being worked on by the
> openEuler community, and the implementation is available in gitee repo [2].
>
> We're going to have a conference call to discuss the potential
> collaborations between CompassCI and LAVA/LKFT. The tech leader of
> CompassCI will do some introduction about CompassCI and present their ideas
> about integrating LAVA/LKFT with CompassCI. Below is the meeting info and
> welcome to join.
>
> Topic: Compass-CI and LAVA/LKFT sync
> Time: Nov 17, 2020 09:00 PM Hong Kong SAR (UTC +8)
>
> Join Zoom Meeting
> https://linaro-org.zoom.com.cn/j/97479537073
>
> Meeting ID: 974 7953 7073
> One tap mobile
> +16699009128,,97479537073# US (San Jose)
> +12532158782,,97479537073# US (Tacoma)
>
> Dial by your location
> +1 669 900 9128 US (San Jose)
> +1 253 215 8782 US (Tacoma)
> +1 301 715 8592 US (Washington D.C)
> +1 312 626 6799 US (Chicago)
> +1 346 248 7799 US (Houston)
> +1 646 558 8656 US (New York)
> 888 788 0099 US Toll-free
> 877 853 5247 US Toll-free
> Meeting ID: 974 7953 7073
> Find your local number: https://linaro-org.zoom.com.cn/u/aFYmaUrqs
>
> [1] https://compass-ci.openeuler.org/
> [2] https://gitee.com/openeuler/compass-ci
>
> Regards,
> Jammy
> --
> Lkq-dev mailing list
> Lkq-dev(a)op-lists.linaro.org
> https://op-lists.linaro.org/mailman/listinfo/lkq-dev
>
Hi All,
CompassCI [1] is a new CI testing framework being worked on by the
openEuler community, and the implementation is available in gitee repo [2].
We're going to have a conference call to discuss the potential
collaborations between CompassCI and LAVA/LKFT. The tech leader of
CompassCI will do some introduction about CompassCI and present their ideas
about integrating LAVA/LKFT with CompassCI. Below is the meeting info and
welcome to join.
Topic: Compass-CI and LAVA/LKFT sync
Time: Nov 17, 2020 09:00 PM Hong Kong SAR (UTC +8)
Join Zoom Meeting
https://linaro-org.zoom.com.cn/j/97479537073
Meeting ID: 974 7953 7073
One tap mobile
+16699009128,,97479537073# US (San Jose)
+12532158782,,97479537073# US (Tacoma)
Dial by your location
+1 669 900 9128 US (San Jose)
+1 253 215 8782 US (Tacoma)
+1 301 715 8592 US (Washington D.C)
+1 312 626 6799 US (Chicago)
+1 346 248 7799 US (Houston)
+1 646 558 8656 US (New York)
888 788 0099 US Toll-free
877 853 5247 US Toll-free
Meeting ID: 974 7953 7073
Find your local number: https://linaro-org.zoom.com.cn/u/aFYmaUrqs
[1] https://compass-ci.openeuler.org/
[2] https://gitee.com/openeuler/compass-ci
Regards,
Jammy
Le jeu. 6 août 2020 à 20:47, Dan Rue via Lkq-dev <
lkq-dev(a)op-lists.linaro.org> a écrit :
> On Fri, Jul 31, 2020 at 11:07:54AM +0000, Milosz Wasilewski via Lkq-dev
> wrote:
> > Hi,
> >
> > RPi4 dispatchers seem to work well in case of LITE devices. I think
> > they might solve some of the problems we have with the current 'big
> > dispatcher' model in LKFT. This however comes with a potential threat.
> > RPi are usine ARM64 architecture which is not officially supported by
> > Android tools (adb, fastboot). They do work so I might just be
> > overreacting, but I would like to know your opinion. What should we do
> > in the LAB
> > a) yes, please try using RPi4 as a dispatcher for Beagle X15
> > b) it's a silly idea and a waste of time. Results from such tests can
> > never be trusted.
>
> (a) - it seems like the approach is viable! My only concerns are
> ensuring that we have multiarch containers for docker, which are still
> seemingly a pain to build under automation.
>
lava-dispatcher and lava-server on docker hub are multiarch containers.
--
Rémi Duraffort
LAVA Architect
Linaro
On Fri, Jul 31, 2020 at 11:07:54AM +0000, Milosz Wasilewski via Lkq-dev wrote:
> Hi,
>
> RPi4 dispatchers seem to work well in case of LITE devices. I think
> they might solve some of the problems we have with the current 'big
> dispatcher' model in LKFT. This however comes with a potential threat.
> RPi are usine ARM64 architecture which is not officially supported by
> Android tools (adb, fastboot). They do work so I might just be
> overreacting, but I would like to know your opinion. What should we do
> in the LAB
> a) yes, please try using RPi4 as a dispatcher for Beagle X15
> b) it's a silly idea and a waste of time. Results from such tests can
> never be trusted.
(a) - it seems like the approach is viable! My only concerns are
ensuring that we have multiarch containers for docker, which are still
seemingly a pain to build under automation.
Dan
>
> milosz
> --
> Lkq-dev mailing list
> Lkq-dev(a)op-lists.linaro.org
> https://op-lists.linaro.org/mailman/listinfo/lkq-dev
--
Linaro LKFT
https://lkft.linaro.org
Hi,
RPi4 dispatchers seem to work well in case of LITE devices. I think
they might solve some of the problems we have with the current 'big
dispatcher' model in LKFT. This however comes with a potential threat.
RPi are usine ARM64 architecture which is not officially supported by
Android tools (adb, fastboot). They do work so I might just be
overreacting, but I would like to know your opinion. What should we do
in the LAB
a) yes, please try using RPi4 as a dispatcher for Beagle X15
b) it's a silly idea and a waste of time. Results from such tests can
never be trusted.
milosz
