+ Milosz
On 28 April 2016 at 13:36, Daniel Thompson daniel.thompson@linaro.org wrote:
On 28/04/16 05:32, Naresh Kamboju wrote:
On 28 April 2016 at 08:40, Leo Yan leo.yan@linaro.org wrote:
On Wed, Apr 27, 2016 at 09:27:37PM +0530, Naresh Kamboju wrote:
[ May be an off topic ]
Hi Leo,
As we discussed offline. I would like to be a part of EAS testing. Since I have Juno board on my desk it would be easy for me to run tests from tools like Workload automation and LISA.
Welcome :)
Milosz suggested to investigate the ways to automate this on LAVA like our team did for Workload automation [1].
If can use automatic method to launch EAS regression testing, this definitely will be very useful and save time.
The steps in Lisa's README.rd: $ source init_env $ nosetests -v tests/eas/rfc.py ./tools/report.py --base noeas --tests eas
Hi Leo,
Just curious, will you plan to enable this into wa2-lava.git?
Yes Because, adb and ssh connection frame work can be re-used from wa2-lava.git
Allowing us to run tests included with lisa (and other lisa derivations) would certainly be useful.
However, in addition, will it be possible for wa2-lava to generate lisa hacking session jobs as well? The automation used to construct KVM images from lisa repositories is useful for interactive work as well.
To explain a bit more...
One of the key features of lisa is that it can be used to interactively explore the effects of different synthetic workloads on a system (for example whilst seeking to reproduce "bad" behavior observed in a real system). Effectively it can act as a tool to help develop complex test cases, especially those that include energy metering.
At this point of time wa2-lava.git for full automation from end-to-end. The design was focused for automation only.
Download-> Configure -> and Install-tests Then running adb or ssh session to connect target device from KVM. Run The tests on target Collect the results on host parse the results on host Attach to LAVA job.
For interactive way of running test is not in the scope now. I am adding Milosz for getting his thoughts on this subject.
- Naresh
Daniel.
- Naresh
[1] https://git.linaro.org/qa/wa2-lava.git
Best regards Naresh Kamboju
On 27 April 2016 at 19:21, Leo Yan leo.yan@linaro.org wrote:
On Mon, Apr 25, 2016 at 03:35:56PM +0100, Neil Williams wrote:
[...]
In each of those logs, I see the smsc911x message:
https://validation.linaro.org/scheduler/job/845552/log_file#L_283_152
https://validation.linaro.org/scheduler/job/845472/log_file#L_349_152
As spotted already, the problem is not that the driver isn't loading, it's that the state is inconsistent:
https://validation.linaro.org/scheduler/job/845552/log_file#L_285_9 Configuring network interfaces... [ 8.449499] smsc911x 18000000.ethernet eth0: SMSC911x/921x identified at 0xffffff8000120000, IRQ: 29285.10 udhcpc (v1.24.1) started 285.11 Sending discover... 285.12 Sending discover... 285.13 Sending select for 10.7.0.20... 285.14 Lease of 10.7.0.20 obtained, lease time 3600 285.15 /etc/udhcpc.d/50default: Adding DNS 10.0.0.2285.16 done.
https://validation.linaro.org/scheduler/job/845472/log_file#L_351_10 Configuring network interfaces... ifup: interface lo already configured 351.11 ifup: interface eth0 already configured 351.12 done.351.13 Starting OpenBSD Secure Shell server: sshd
After I chanaged the command to "ifup -a -f" in rootfs's file /etc/init.d/networking, then I can boot up 4 times with multinode definition, everytime Juno can get dynamic IP for Juno board successfully.
So this issue is caused by mismatch info in /etc/network/run/ifstate, is this caused by LAVA boots Juno boards for several times during this job and one previous boot time may write dirty info into ifstate?
Thanks, Leo Yan _______________________________________________ Lava-users mailing list Lava-users@lists.linaro.org https://lists.linaro.org/mailman/listinfo/lava-users