- lava-users - op-lists.linaro.org

by Li, Xuelian

Hi，Neil： I want to control device power on/off by lavapdu.,but when I run pduclient --daemon autotest001 --hostname 10.5.142.6 --command off --port 10 ERROR : Error sending command, wrong daemon hostname? Can you help me solve this problem? Lava-server :IP 10.5.21.5 autotest001 Pdudaemon run on the lava-server Lavapdu.conf: { "daemon": { "hostname": "10.5.21.5", "port": 16421, "dbhost": "127.0.0.1", "dbuser": "pdudaemon", "dbpass": "pdudaemon", "dbname": "lavapdu", "retries": 5, "logging_level": "INFO" }, "pdus": { "192.168.10.2": { "driver": "apc9210" }, "192.168.10.3": { "driver": "apc7952", "telnetport": 5023 }, "192.168.10.4": { "driver": "apc7952" }, "192.168.10.5": { "driver": "apc8959", "this": "that", "something": "else" }, "192.168.25.52": { "driver": "apc7952" } } This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years

2
1
0 0

auth token

by Xu, Hongyu

Hi lava experts: I deploy the lava 2017.7 with SSL Self-signed certificate, then I want to create the device-dictionary for the qemu devices, it’s error when I add auth_token. Please give me some suggest! lava-tool auth-add --token-file /home/lavaadmin/xuhy/qemu_test_token --user-shortcut 1680141 --endpoint-shortcut qemu_test01 https://1680141@10.5.21.5 ERROR: <urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:661)> Best Regards XuHongyu From: Xu, Hongyu Sent: 2017年9月4日 14:41 To: 'Lava-users(a)lists.linaro.org' <Lava-users(a)lists.linaro.org> Subject: Local Django user accounts Hi all, I have installed the lava-serer (2017.7) on my DEBIAN 9.10. But now I can’t sign in with the local Django user, this is the error infomations: Forbidden (403) CSRF verification failed. Request aborted. You are seeing this message because this site requires a CSRF cookie when submitting forms. This cookie is required for security reasons, to ensure that your browser is not being hijacked by third parties. If you have configured your browser to disable cookies, please re-enable them, at least for this site, or for 'same-origin' requests. More information is available with DEBUG=True. Please give me some advices! root@hxtcorp:~# dpkg-query -l lava* Desired=Unknown/Install/Remove/Purge/Hold | Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend |/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad) ||/ Name Version Architecture Description +++-=================================-=====================-=====================-======================================================================= ii lava-coordinator 0.1.7-1 all LAVA Coordinator daemon un lava-dashboard <none> <none> (no description available) un lava-dashboard-tool <none> <none> (no description available) ii lava-dispatcher 2017.7-1~bpo9+1 amd64 Linaro Automated Validation Architecture dispatcher un lava-scheduler <none> <none> (no description available) un lava-scheduler-tool <none> <none> (no description available) ii lava-server 2017.7-1~bpo9+1 all Linaro Automated Validation Architecture server ii lava-server-doc 2017.7-1~bpo9+1 all Linaro Automated Validation Architecture documentation ii lava-tool 0.21-1 all command line utility for LAVA un lavapdu <none> <none> (no description available) ii lavapdu-client 0.0.5-1 all LAVA PDU client ii lavapdu-daemon 0.0.5-1 all LAVA PDU control daemon add the local Django user: sudo lava-server manage createsuperuser --username $USERNAME --email=$EMAIL sudo lava-server manage authorize_superuser --username $USERNAME root@hxtcorp:/etc/lava-server# lava-server manage users list List of users: * 1680077 * 1680141 * lava-health * xuhy Best Regards XuHongyu This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years

2
2
0 0

adding a device dictionary in 2017.7

by Daniel Sangorrin

Hi, I upgraded to LAVA 2017.7 (for now I'm using https://github.com/kernelci/lava-docker) and managed to add my device-type and device as follows: # Note: I didn't know how to update to 2017.9 docker# lava-server manage device-types add renesas-iwg20m docker# lava-server manage devices add --device-type renesas-iwg20m --worker lava-docker iwg20m01 However, when I want to add the device dictionary I get an error. # Note: I am following https://validation.linaro.org/static/docs/v2/pipeline-admin.html First, I tried with "lava-server manage device-dictionary" as in the documentation, but I failed: docker# lava-server manage device-dictionary Unknown command: 'device-dictionary' Is this only available in the 2017.9 version? Then, I tried with " lava-tool device-dictionary localhost lava-docker --update myiwg20m.jinja2" but I get: Updating device dictionary for lava-docker on localhost ERROR: Endpoint URL not found in the authorization list: localhost [Note] my lava-tool is version 0.23-1 (the newest according to apt) Now, there is a section saying "Using the command line The device dictionary exists as a jinja2 file in /etc/lava-server/dispatcher-config/devices and can be updated by admins with the necessary access.". Is this equivalent to using lava-server o lava-tool to install a device dictionary?? Others: - I confirmed that Pipeline device? is checked. - Django administration > Devices: what is the meaning of Hostname? I have a value of iwg20m01 (the name of the device), there is no relation with the server/container's hostname, is there? - There is an authentication token for the user 'kernel-ci' Thanks, Daniel

8 years

2
2
0 0

Manual testing in Lava

by Denis HUMEAU

Hello, My question looks pretty silly, but as Lava has a big set of features that I haven't discovered yet, I dare ask. Lava is designed for automatic testing. But everything cannot be fully automated, at least in a reasonable development time frame. More precisely, tests execution can be automated, but the test result needs to be given by a tester. This will happen for example in display or graphics tests. So, is there a simple way in Lava to let a tester give a manual test result, during or after the test execution? I already have in mind pretty complicated solutions based on Multinode, but simple is key here Thanks, Denis

8 years

2
1
0 0

problems with initramfs in LAVA 2017-7

by Robert Marshall

Hi I have a locally built initramfs which worked with a beaglebone black health check in LAVA 2016-12 but with version 2017-7 the health check is failing. I've tried using an initramfs from http://images.validation.linaro.org/snapshots.linaro.org/components/lava/st… with 2017-7 and that works ok. The various sections of the yaml file are as below and a comparison of the 2 boots are at http://paste.baserock.org/iyuzakogez , a lot more noise from the local one but 2016-12 was happy with that. Is there something I'm missing with the update that I need to change in this case? ramdisk: # url: http://images.validation.linaro.org/snapshots.linaro.org/components/lava/st… # locally built version url: http://localhost:8010/initramfs/initramfs-local.cpio.gz compression: gz add-header: u-boot .... # BOOT_BLOCK - boot: method: u-boot commands: ramdisk prompts: - '$initramfs$' timeout: minutes: 2

8 years

2
4
0 0

Switching LAVA to V2

by Dave Pigott

Hi everyone, As many of you may know, LAVA V2 has been in development and is now ready to roll out across all instances, and we have started this transition with most micro-instances now being predominantly V2. That leaves the main production instance, validation.linaro.org. Of the supported device types in V2 16 of the 21 have been fully switched over already. Most of these devices can accept V1 and V2 job submissions, but for underlying structural reasons some device types can *only* be V1 or V2, not both. This is to let you know that in the week following SFO17 we will be switching those devices (HiKey, Drangboard 410c, X15, IFC6410 and Mustang) to V2 only, and any job submissions that are in the old JSON format will be rejected by the LAVA scheduler. If you have bots that submit V1 jobs to those device types, or if you manually submit V1 jobs to those device types, you are strongly urged to switch those off and move to V2 submissions. The next release of LAVA software will not support V1 at all, so for anyone submitting V1 jobs to *any* device, you will need to move to V2 by the beginning of November at the latest. Many thanks Dave ---------------- Dave Pigott LAVA Lab Lead Linaro Ltd t: (+44) (0) 1223 400063

8 years

1
0
0 0

Start a flashed board and launch tests

by Philippe BEGNIC

Hello Lava team, I am trying to test a flashed board that does'nt require to upload tests images ( uImage dtb, rootfs). The images have been previously flashed on the board under test, by an external Flasher. To do that, I am using the boot method "minimal" that we have been using for other cases. In my test, I power on the board, and wait for the kernel prompt ("root@"). When I got the prompt, LAVA try to mount the test partition (previously prepared in deploy phase), but it fails. I get the following message "No connection retrieved from namespace data". * Is it possible to do this operation with the boot method "minimal" ? * Do you have jobs example that do the same operations ? Find below the boot and test phase I am using : ( In the deploy phase, I have downloaded images that I don't use. ) - boot: method: minimal failure_retry: 2 prompts: - 'root@' timeout: minutes: 2 - test: timeout: minutes: 5 namespace: flash_to_sd definitions: - from: git repository: /local/dev/frq93151/dispatcher/lava-tests path: tests-def/ddr_test_definition.yaml name: ddr_tests BR Philippe

8 years

1
0
0 0

Local Django user accounts

by Xu, Hongyu

Hi all, I have installed the lava-serer (2017.7) on my DEBIAN 9.10. But now I can't sign in with the local Django user, this is the error infomations: Forbidden (403) CSRF verification failed. Request aborted. You are seeing this message because this site requires a CSRF cookie when submitting forms. This cookie is required for security reasons, to ensure that your browser is not being hijacked by third parties. If you have configured your browser to disable cookies, please re-enable them, at least for this site, or for 'same-origin' requests. More information is available with DEBUG=True. Please give me some advices! root@hxtcorp:~# dpkg-query -l lava* Desired=Unknown/Install/Remove/Purge/Hold | Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend |/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad) ||/ Name Version Architecture Description +++-=================================-=====================-=====================-======================================================================= ii lava-coordinator 0.1.7-1 all LAVA Coordinator daemon un lava-dashboard <none> <none> (no description available) un lava-dashboard-tool <none> <none> (no description available) ii lava-dispatcher 2017.7-1~bpo9+1 amd64 Linaro Automated Validation Architecture dispatcher un lava-scheduler <none> <none> (no description available) un lava-scheduler-tool <none> <none> (no description available) ii lava-server 2017.7-1~bpo9+1 all Linaro Automated Validation Architecture server ii lava-server-doc 2017.7-1~bpo9+1 all Linaro Automated Validation Architecture documentation ii lava-tool 0.21-1 all command line utility for LAVA un lavapdu <none> <none> (no description available) ii lavapdu-client 0.0.5-1 all LAVA PDU client ii lavapdu-daemon 0.0.5-1 all LAVA PDU control daemon add the local Django user: sudo lava-server manage createsuperuser --username $USERNAME --email=$EMAIL sudo lava-server manage authorize_superuser --username $USERNAME root@hxtcorp:/etc/lava-server# lava-server manage users list List of users: * 1680077 * 1680141 * lava-health * xuhy Best Regards XuHongyu This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years

2
1
0 0

help:Adding pipeline workers to the master

by kingboli

Hi all, I used LAVA V1 before and I want to migrate to LAVA V2 following the guide " First steps installing LAVA V2" . I had a problem when adding pipeline workers to the master. After input the command " sudo lava-server manage workers add <HOSTNAME>", I got error "Unknown command: 'workers'" and I failed to find help by typing "lava-server -h manage". Can anybody help me on this step? BTW, I want to know whether multiple workers can have the same hostname ? Thanks in advance.

8 years

2
5
0 0

Licenses in test-definitions

by Robert Marshall

Hi, The tests in https://git.linaro.org/qa/test-definitions.git in some cases have license in the subdirectories but there's no top level licence file covering the whole repository. Would you like a PR submitting to cover this or do you envisage the maintainers making this change themselves? We're hoping to use some of these and so wanted licences to be consistent. We'd also be prepared to review licences in future repository changes. The licenses that I've found all seem to be GPL v2 so I assume that is the license flavour which should be covering the whole repos. Robert

8 years

4
5
0 0

[results error]constants section not present in the device config.

by Michael Chen

Hi Lava users, I just updated lava from 2016.12 to 2017.6-1. And I found my health-check job failed(uncomplete) after the update. Here is the result logs before: https://pastebin.com/KWAnYQvD and after: https://pastebin.com/mqXDfR3H I checked the log and think the cause for the issue is: ``` case: job definition: lava error_msg: constants section not present in the device config. error_type: Configuration result: fail ``` What I have done: * Try to find the "constants section" with no luck. * Looked at the logs in /var/lava-server/ and didn't find the log specific for the reason of this. * Asked IRC channel and ask, answered by codehelp(thanks:)). Then tried to replace my device-type config in /etc/lava-server/dispatcher-config/device-types with x86.jinja2 and the error message is the same. Any input would by much appreciated, thank you. Regards, -Michael Chen

8 years

2
2
0 0

Lava debian stretch repo

by Sharma, Varun

Hi Lava Team, I am using Lava image on debian stretch from https://images.validation.linaro.org/production-repo stretch-backports main & in It chart/query api is working. Then I took latest code from lava release branch https://github.com/Linaro/lava-server/commits/release to my local branch Tag taken :- c0b431f1b56227f6b97ba0e7859bfcc3fb29b62f refs/tags/2017.7 & build it but Lava charts is not coming when I run lava server["python manage.py runserver"] using latest code from git?? Can you please share where is latest code being used in debian stretch located? Thanks, Varun

8 years

2
1
0 0

Is it possible to request a device by using device name

by Arthur She

Hi, In LAVA v1, as I remember, I can use either "device" or "device_type" to request a device for the test. If I use "device", for example "device": "beaglebone-black03", then the specific device, beaglebone-black03, will be used for the test. Can I do the same thing in LAVA v2? To request a device by using device name? Thanks, Arthur

8 years

3
9
0 0

Jobs stuck in canceled mode and blocking queue

by Robert Marshall

Hi, I've got 2 jobs stuck in canceled mode which are preventing any other job from running. I'm running lava (2017-7) in a VM and have tried rebooting the VM to clear the issue but without success (ie the jobs still block the queue). an extract from /var/log/lava-server/django.log is attached I get this 500 error when viewing the results for the job Is there a manual way of clearing this? The health check has notification associated with it (and set to verbose) and every time I reboot I get an email and irc saying that it's finished! Robert

8 years, 1 month

2
6
0 0

NFSroot problems

by Benjamin Auclair

Hi everyone, I'm implementing the nfsroot for my devices but it seems that when extracting my rootfs.tar.xz lava keep the parent folder rootfs and didn't extract all the files in extract-nfsrootfs-XXXXX so that the lava test overlay is putting outside of rootfs folder and raises an error during execution. Is there a way in job definition to ask lava to put the test overlay in the rootfs folder ? Thanks Benjamin AUCLAIR -- - Confidential -

8 years, 1 month

2
1
0 0

Stuck in lava-test-shell

by Benjamin Auclair

Hi everyone, hope you're fine ! I'm quite stuck in my platform development: indeed, I succedded in adding my own device type, I'm able to boot on linux by TFTP and to perform auto-login actions. However, I face difficulties with test-shell. I have the following error: https://pastebin.com/grPcvb14 And the definition of stage is: https://pastebin.com/ArV11Gbb Stage value seems to be none and I also realized that my test shell isn't downloaded from git during server processing. Thus I think, that stage is empty because the test shell definition isn't in the temporary files. Am I wrong ? Even if I think I found my problem, I don't know how to solve it, may it be due to my device-type config ? Thanks a lot and have a nice day ! Benjamin AUCLAIR -- - Confidential -

8 years, 1 month

3
5
0 0

Could not set UBOOT_AUTOLOAD_MESSAGE variable while Adding BBB on LAVA server

by ankit gupta

Hi All, I am adding BBB board on LAVA server I want to change "UBOOT_AUTOLOAD_MESSAGE" in constant.py, I used "interrupt_prompt" parameters in job submission but it took the message written in constant.py. If I changed the message in constant.py its working but I know this is not the right way to do that, Please suggest if any one has idea what is the problem with me. Below is the my Job: device_type: beaglebone-black # NFS fails on panda and arndale. job_name: BBB smoke test timeouts: job: minutes: 240 action: minutes: 240 connection: minutes: 2 priority: medium visibility: public metadata: source: https://git.linaro.org/lava-team/refactoring.git path: health-checks/beaglebone-black-health.yaml build-readme: http://images.validation.linaro.org/snapshots.linaro.org/components/lava/st… build-console: https://ci.linaro.org/view/lava-ci/job/lava-debian-armmp-armhf/1/console build-script: http://images.validation.linaro.org/snapshots.linaro.org/components/lava/st… actions: - deploy: timeout: minutes: 40 to: tftp kernel: url: file:////home/pi/lava/dl/vmlinuz ramdisk: url: file:////home/pi/lava/dl/initramfs.cpio.gz compression: gz # the bootloader needs a u-boot header on the modified ramdisk add-header: u-boot modules: url: file:////home/pi/lava/dl/modules.tar.gz compression: gz nfsrootfs: url: file:////home/pi/lava/dl/jessie-armhf-nfs.tar.gz compression: gz os: debian dtb: url: file:////home/pi/lava/dl/am335x-boneblack.dtb - boot: method: u-boot commands: nfs parameters: shutdown-message: 'INIT: Sending processes the TERM signal' interrupt_prompt: 'Press SPACE to abort autoboot in 10 seconds' interrupt_char: ' ' send_char: False type: bootz auto_login: login_prompt: 'beaglebone login: ' username: root prompts: - 'root@jessie:' timeout: minutes: 10 - test: timeout: minutes: 50 definitions: - repository: git://git.linaro.org/qa/test-definitions.git from: git path: ubuntu/smoke-tests-basic.yaml name: smoke-tests

8 years, 1 month

3
6
0 0

mount worker host directory to LXC instance

by Chetan Sharma

Hi Lava Team I am trying to mount a directory of worker (host computer) to LXC instance running on it with job. I have added this entry in lxc default configuration file of worker computer. Default Configuration file path is : /etc/lxc/default.conf Entry is :- lxc.mount.entry = /var/lib/nfsdir var/nfsmnt none bind,create=dir 0 0 After restart lxc service. I executed lava job then this directory "/var/lib/nfsdir" of worker machine is not mounting on LXC instance in directory "/var/nfsmnt". But if i manually create and start lxc instance on worker, then this directory "/var/lib/nfsdir" of worker machine is mounted on LXC instance. Can anyone assist me on this issue, that how i can resolve this issue. -- Thanks & Regards Chetan Sharma

8 years, 1 month

2
3
0 0

LXC share data with DUT

by Chetan Sharma

Hi Lava Team Can you assist me on this usecase that how i can share LXC data with DUT. We have LXC and DUT TEST defined in a Job. LXC Tests generates some data and logs which is required to be accessed by DUT TEST. -- Thanks & Regards Chetan Sharma

8 years, 1 month

4
3
0 0

Installing LAVA on Stretch

by Neil Williams

There are delays getting packages into stretch-backports, as expected. In the meantime, this is a reminder of how to use backports: first start with stable. So when installing LAVA on Stretch, even if what you want is the latest release from production-repo or staging-repo (currently 2017.7), then your first step is to install 2016.12 from Stretch. # apt -y install lava-server # apt -y install vim # wget http://images.validation.linaro.org/production-repo/production-repo.key.asc # apt-key add production-repo.key.asc # vim /etc/apt/sources.list.d/lava.list # # edit the file to specify: deb http://images.validation.linaro.org/production-repo jessie-backports main Yes, that is jessie-backports - we don't have packages in stretch-backports at this time. # apt update # apt upgrade If you specify backports too early, you'll get a range of other packages from backports - you don't actually need jessie-backports or stretch-backports from Debian at this time. Packages for jessie-backports are built on jessie. Packages for stretch-backports are built on stretch. It's the same source code in each case. Right now, there aren't any problems with installing from jessie-backports on stretch - particularly if you install lava-server from stretch in the first place so that the majority of your packages come from stretch. -- Neil Williams ============= neil.williams(a)linaro.org http://www.linux.codehelp.co.uk/

8 years, 1 month

2
2
0 0

Facing issue while installing LAVA 2016.6

by ankit gupta

Hello LAVA Users, I want to install LAVA version 2017.6 or 2017.7 on debian-9 (Stretch), I followed the instruction given on https://staging.validation.linaro.org/static/docs/v2/installing_on_debian.h… For this I added production repo in my *sources.list* file, the content of my sources.list is given below. *deb http://ftp.us.debian.org/debian/ <http://ftp.us.debian.org/debian/> stretch main* *deb-src http://ftp.us.debian.org/debian/ <http://ftp.us.debian.org/debian/> stretch main* *deb http://security.debian.org/debian-security <http://security.debian.org/debian-security> stretch/updates main* *deb-src http://security.debian.org/debian-security <http://security.debian.org/debian-security> stretch/updates main* *deb http://ftp.us.debian.org/debian/ <http://ftp.us.debian.org/debian/> stretch-updates main* *deb-src http://ftp.us.debian.org/debian/ <http://ftp.us.debian.org/debian/> stretch-updates main* *deb http://ftp.fr.debian.org/debian <http://ftp.fr.debian.org/debian> stretch-backports main* *deb https://images.validation.linaro.org/production-repo <https://images.validation.linaro.org/production-repo> sid main* After this I run *"apt-get update" & "apt-get install lava-server"*, This command is given me error as given below. *root@debian:/home/debian# apt-get install lava-server* *Reading package lists... Done* *Building dependency tree * *Reading state information... Done* *Some packages could not be installed. This may mean that you have* *requested an impossible situation or if you are using the unstable* *distribution that some required packages have not yet been created* *or been moved out of Incoming.* *The following information may help to resolve the situation:* *The following packages have unmet dependencies:* * lava-server : Depends: lava-dispatcher (>= 2017.5) but it is not going to be installed* * Recommends: lava-coordinator but it is not going to be installed* * Recommends: lava-server-doc but it is not going to be installed* *E: Unable to correct problems, you have held broken packages.* I tried *staging repo* as well and also *jassie-backports* but couldn't proceed, Please anyone have some ide what going wrong in my case and how I can resolve this issue. My main objective is to install *LAVA 2017.6/2017.7* on my debian stretch. Thanks, Ankit

8 years, 1 month

2
1
0 0

Enable interface in LXC to access internet inside LXC

by Chetan Sharma

Hi Lava Team, Can anyone assist me with these 2 quires as 1. Can we enable interface inside LXC to access host network, so that we can access any device on network of host machine and can access internet inside LXC to execute script. 2. Can we pass Params in Lava job, which can propagate to lava-test-action job or yaml. If possible can you guide me with process to perform this action ? share a reference job which perform this task. -- Thanks & Regards Chetan Sharma

8 years, 1 month

2
1
0 0

Lava V2: Multinode with Linux PC?

by Denis HUMEAU

Dear all, Do you have any example or recommendation on the way to implement the following use case in Lava? DUT under Linux A PC under Linux to use as a "server" for VLC or iPerf test cases I guess Multinode can be used to manage these use cases, but I don't know how to deal with the constraints linked to the PC: no reboot, no dedicated test kernel. Thanks, Denis

8 years, 1 month

3
2
0 0

Re: [Lava-users] how to start QEMU environment variables

by Jan-Simon Möller

Hi Neil, Kevin ! > On 20 July 2017 at 02:57, Kevin Hilman <khilman at baylibre.com> wrote: > > Hello, > > > > There are many configurable options when starting QEMU that are > > controlled by environment variables, and I'm trying to figure out how to > > pursuade LAVA v2 to set them. > > It's not something which should be done with the QEMU device type > because that is installed on the worker - it is not a device that can > be reconfigured to suit test writers and much of the functionality of > QEMU must remain off-limits when QEMU is running on the worker. > > > As an example, configuring the host audio driver used by QEMU is set by > > running QEMU with: QEMU_AUDIO_DRV=<option> qemu-system-x86_64. > > The worker is typically a server, it might not have any audio support. > Even if it does, that functionality is off-limits to test writers. The > available options to QEMU must be tightly constrained to protect other > tasks on the worker. A see. There is a valid point here in that there should not be control of the environment from the 'outside' though a job description. > > > Unfortunately, device-types/qemu.jinja2 doesn't provide anyway to > > override the qemu-system binary used (it's conditional based on arch), > > This is deliberate, to protect the worker from aberrant processes and > to give the admins sufficient control over how the worker behaves. Now for the above case: QEMU_AUDIO_DRV=<option> qemu-system-* I see this as a *very valid env var* for a server-side deployment - exactly as you said - a server does not have a sound card and QEMU_AUDIO_DRV=none is what we need to use there to prevent qemu from looking for a host sound card. We should even set this by default - we don't expect any sound to reach the worker - do we ;) ? I don't think so. IMHO this should be an option set on the worker node in the lava- dispatcher.conf (e.g. LAVA_EXTRA_QEMU_ENV="QEMU_AUDIO_DRV=none"). Enable the admins to choose this. > > No. This is not something that the worker can support. It needs to > happen only on a test device. The worker is not itself a test device. As I said, I think this belongs to the worker setup and we should enable the admins to make their choice. Dipl.-Ing. Jan-Simon Möller jansimon.moeller(a)gmx.de

8 years, 1 month

3
6
0 0

how to start QEMU environment variables

by Kevin Hilman

Hello, There are many configurable options when starting QEMU that are controlled by environment variables, and I'm trying to figure out how to pursuade LAVA v2 to set them. As an example, configuring the host audio driver used by QEMU is set by running QEMU with: QEMU_AUDIO_DRV=<option> qemu-system-x86_64. Unfortunately, device-types/qemu.jinja2 doesn't provide anyway to override the qemu-system binary used (it's conditional based on arch), but even a quick hack to allow it to be overriden[1], and adding the env as a prefix didn't because LAVA assumes the first item is an actual binary expected in $PATH. My attempt led to: Invalid job definition Cannot find command 'QEMU_AUDIO_DRV=none qemu-system-x86_64' in $PATH Seems like there should be a more general way to pass enviornment variables to QEMU that I must be missing. If there's not, would be the recommended way to add this feature? Kevin [1] diff --git a/lava_scheduler_app/tests/device-types/qemu.jinja2 b/lava_scheduler_app/tests/device-types/qemu.jinja2 index 786f53bdb30d..e7c265a3048b 100644 --- a/lava_scheduler_app/tests/device-types/qemu.jinja2 +++ b/lava_scheduler_app/tests/device-types/qemu.jinja2 @@ -41,7 +41,7 @@ actions: {% elif arch == 'arm' %} qemu-system-arm {% elif arch == 'amd64' or arch == 'x86_64' %} - qemu-system-x86_64 + {{ qemu_command|default('qemu-system-x86_64') }} {% elif arch == 'i386' %} qemu-system-i386 {% endif %} With that change, I added context: qemu_command: "QEMU_AUDIO_DRV=none qemu-system-x86_64" to the job definition

8 years, 1 month

2
1
0 0

How to expose device path like '/dev/ttyUSB0' to LAVA-LXC?

by Li, Yifan2

Hello LAVA experts, Is it possible to expose a device path like '/dev/ttyUSB0' to lxc in job definition or device definition? I have a feature requirement to control a relay card inside LAVA lxc. I have got clues for device_info and read https://validation.linaro.org/static/docs/v2/lava-scheduler-device-dictiona… and https://validation.linaro.org/static/docs/v2/admin-lxc-deploy.html?highligh…. However I did not know how to use it to expose '/dev/ttyUSB0' to LXC. Thanks and Best Regards, Yifan

8 years, 1 month

2
3
0 0

Regarding Image Charts2.0

by Sharma, Varun

Dear lava Community, I want to use Image charts2.0 for viewing my lava job results. I am using "lava-test-case" to check pass/fail & I am getting results also. steps: - lava-test-case linux-linaro-ubuntu-pwd --shell pwd - lava-test-case linux-linaro-ubuntu-uname --shell uname -a - lava-test-case linux-linaro-ubuntu-vmstat --shell vmstat I want to know how to get these results in image charts, I can see it ask to add chart & addfilter , but there no data is available when I try to add filter? Similarly if I have to use query api, what kind of query should be used to fetch test data from lava suite? Detailed info to use image chart will be appreciated, as I am new to Charts/Lava...or at any link. Thanks, Varun

8 years, 1 month

3
10
0 0

Help adding new device type

by Benjamin Auclair

Hi everyone, I'm trying to add own device-type to my lab, but I'm facing some difficulties when running jobs. I have the following log error: https://pastebin.com/Eznq6RLA I clearly understand the log but I'm not able to figure out what to do: I thought it will be enough describing boot/power actions into device-type. But it seems not... Do I need to create a .conf file into /usr/lib/python2.7/dist-packages/lava_dispatcher/default-config/lava-dispatcher/device-types folder ? By the way I'm not sure to understand the .conf purpose ? Are they here, only to be some default files ? I attached my device-type and my job, maybe it will help you ! Thanks a lot ;) P.S: I already did some jobs on Qemu and bbb and read the whole documentation. -- - Confidential -

8 years, 1 month

2
2
0 0

multinode definition parsing error

by Milosz Wasilewski

Hi, I'm not entirely sure this job definition is correct, but the only error I'm getting is only "Problem with submitted job data: Invalid YAML - check for consistent use of whitespace indents.". The YAML validates just fine so I'm unsure what is wrong. Any hints? The YAML in question is enclosed. milosz

8 years, 1 month

2
1
0 0

Lava V2 via command line

by Denis HUMEAU

Hello, We use a python script, LAVA_via_API, to trigger our test jobs. I must say that I don't know whether this script is a pure internal creation or whether it's been inspired by a Linaro script. Its role is simple, create a lava job with quite a few parameters (job name, server, worker, kernel, rootfs, dtb, device, device_type, and so on), submit the job, get results and logs. Whatever, before reworking completely this script, I assume that a reference one exists on one of the Linaro gits. Can you tell me where to find this? Thanks, Denis

8 years, 1 month

2
1
0 0

Re: [Lava-users] Seek Assistance: LAVA-lxc Job creation and submition

by Neil Williams

ALWAYS keep the list in CC please. On 7 July 2017 at 10:28, Chetan Sharma <chetan.sharma(a)timesys.com> wrote: > Hi Neil > Thanks for sharing detailed information to work with LXC. > 1. I am following sample pipeline job with BBB. > https://git.linaro.org/lava/lava-dispatcher.git/tree/lava_dispatcher/pipeli… > > I have modified this job details with following value, but i am getting an > Error as ['Invalid job - missing protocol'] Check with the valid job: https://staging.validation.linaro.org/scheduler/job/178130 > As i have defined protocol as "lava-lxc" which is valid protocol, But job > object does not have any protocol details, i have verified by printing > details of self.job.protocols is [] Then that is an invalid job. Your modifications have got something wrong. There are a lot of changes in your file against the example. Change only one thing at a time. > 2. How test action execute on lxc and device ? Run a test action in the same namespace as the LXC. In the case of the example, namespace: probe. https://staging.validation.linaro.org/scheduler/job/178130/definition#defli… > Can we execute test action in this process > First lxc test action execute ---> Device1 test action execution start > -> Device2 test action execution start > > > ================================================== > device_type: beaglebone-black > > job_name: LXC and bbb Please attach files to emails to the list. There's no need to quote the entire file to the list. Take time to understand namespaces. The LXC is transparent and the namespaces are used to tie the different actions together into one sequence in the LXC and one sequence for the test device. LXC protocol support is not the same as MultiNode - operations happen in series. The LXC protocol is not a substitute for MultiNode either. If you need parallel execution, you have to use MultiNode. Split up your test shell definitions if necessary. Also, attach (not include) the full test job log because that contains details of the package versions being used and other information. > On Fri, Jul 7, 2017 at 1:32 PM, Neil Williams <neil.williams(a)linaro.org> > wrote: >> >> On 7 July 2017 at 07:23, Chetan Sharma <chetan.sharma(a)timesys.com> wrote: >> > Hi Everyone >> > Hopefully everyone is doing well in this group. >> > Main intend of writing this email to seek assistance in regard to >> > one >> > feature of Lava as lava-lxc which help to create a LXC instance on >> > worker >> >> >> https://staging.validation.linaro.org/scheduler/job/174215/definition#defli… >> >> > and then we can execute any script on worker and propagate its >> > characteristics and result in another actions of Job on board. >> >> That isn't entirely clear. What are you trying to do in the LXC? >> >> You need to deploy an LXC, start it with a boot action for the LXC and >> then start a test shell in the LXC where you can install the tools you >> need to execute. >> >> Talking to the device from the LXC can be difficult depending on how >> the device is configured. To use networking, you would have to have a >> fixed IP for the device. To use USB, you need to use the device_info >> support in the device dictionary to add the USB device to the LXC. >> >> > I have gone through documentation shared on Lava V2 instance for >> > LXC >> > job creation, but i am not able to successfully execute job on Debian >> > Jessie >> > Release. >> > >> > https://validation.linaro.org/static/docs/v2/deploy-lxc.html#index-0 >> > >> > Can you assist me and share a reference process/document to proceed >> > further to create a job using this feature. >> >> More information is needed on exactly what you are trying to do, how >> the LXC is to connect to the device and what support the device offers >> to allow for those operations. >> >> -- >> >> Neil Williams >> ============= >> neil.williams(a)linaro.org >> http://www.linux.codehelp.co.uk/ > > > > > -- > Thanks & Regards > Chetan Sharma -- Neil Williams ============= neil.williams(a)linaro.org http://www.linux.codehelp.co.uk/

8 years, 2 months

1
0
0 0

Seek Assistance: LAVA-lxc Job creation and submition

by Chetan Sharma

Hi Everyone Hopefully everyone is doing well in this group. Main intend of writing this email to seek assistance in regard to one feature of Lava as lava-lxc which help to create a LXC instance on worker and then we can execute any script on worker and propagate its characteristics and result in another actions of Job on board. I have gone through documentation shared on Lava V2 instance for LXC job creation, but i am not able to successfully execute job on Debian Jessie Release. https://validation.linaro.org/static/docs/v2/deploy-lxc.html#index-0 Can you assist me and share a reference process/document to proceed further to create a job using this feature. Look forward to hear with positive response. -- Thanks & Regards Chetan Sharma

8 years, 2 months

2
1
0 0

New LAVA user looking for help with Yocto with grub

by Elric Hindy

Hi, I have been investigating LAVA for use in our organisation and i'm stuck trying to get a hello world test case running on our hardware and looking for some help. We looked at the YOCTO test tools however it can only use devices with a fixed ip which we can't guarantee or want during our testing as we also test network settings. It's also limited in configuration, LAVA package seems to meet all our requirements however i'm still unsure on how to do a few things. We use Yotco images and boot with the grub bootloader. All our devices are connected via Ethernet only and power and peripheral switching is controlled via usb relays. After reading through all the documentation i'm still unsure of how to set up and actually run a test on our hardware. What tools do i need to install in the test image and how do i get it to communicate with grub? I assume a base image is one that includes nothing but the tools and grub. We have a recovery partition with tiny core which could facilitate that but it's not required for the automated testing. I've used the akbennet/lava-server docker image and it is up and running, although test jobs are scheduled but never run on the qemu devices so a little stuck there. Basically, I need help to get LAVA to connect to one of our devices to load the image and run tests? Choosing the image, writing tests and mostly configuring the pipeline I understand. After 2 weeks i'm posting here hoping someone can assist me. Regards, Elric Elric Hindy Test Engineer T +61 2 6103 4700 M +61 413 224 841 E elric.hindy(a)seeingmachines.com W www.seeingmachines.com<http://www.seeingmachines.com> [Seeing Machines]<https://www.seeingmachines.com/> This email is confidential. If you are not the intended recipient, destroy all copies and do not disclose or use the information within. No warranties are given that this email does not contain viruses or harmful code.

8 years, 2 months

2
1
0 0

Issue in adding lava-worker using R-Pi

by ankit gupta

Hi All, I am trying to setup a remote lab using Raspberry pi on my local network. I installed lava-server and a worker on my laptop and its working fine. I installed raspbian on R-Pi and follow the instruction given on lava site, but when slave is trying to connect to master its not getting any response, I am able to ping master from my R-pi board and default port 3079 is open on my machine. I used no encryption and use URL to connect master as follow. MASTER_URL="tcp://10.42.0.24:3079" LOGGER_URL="tcp://10.42.0.24:3079" I continuosly getting log messgaes like, DEBUG Sending new HELLO_RETRY message to the master (are they both running the same version?) INFO Waiting for the master to reply DEBUG Sending new HELLO_RETRY message to the master (are they both running the same version?) INFO Waiting for the master to reply DEBUG Sending new HELLO_RETRY message to the master (are they both running the same version?) Please, if any one have some idea why I am not able to connect please help. Thanks, Ankit

8 years, 2 months

2
1
0 0

Re: [Lava-users] [Linaro-validation] Further steps with Zephyr tests

by Neil Williams

On Mon, 3 Jul 2017 23:50:25 +0300 Paul Sokolovsky <paul.sokolovsky(a)linaro.org> wrote: > Hello Milosz, > > I appreciate getting at least some response ;-). Some questions > however could use a reply from LAVA team, I guess. > > On Mon, 3 Jul 2017 13:34:49 +0100 > Milosz Wasilewski <milosz.wasilewski(a)linaro.org> wrote: > > [] > > > > jobs submit a number of tests to LAVA (via > > > https://qa-reports.linaro.org/) for the following boards: > > > arduino_101, frdm_k64f, frdm_kw41z, qemu_cortex_m3. Here's an > > > example of cumulative test report for these platforms: > > > https://qa-reports.linaro.org/lite/zephyr-upstream/tests/ > > > > > > That's really great! (Though the list of tests to run in LAVA > > > seems to be hardcoded: > > > https://git.linaro.org/ci/job/configs.git/tree/zephyr-upstream/submit_for_t…) > > > > > > > It is, as I wasn't really sure what to test. The build job needs to > > prepare the test templates to be submitted to LAVA. In case of > > zephyr each tests is a separate binary. So we end up with the > > number of file paths to substitute in the template. Hardcoding was > > the easiest thing to get things running. But I see no reason why it > > wouldn't be changed with some smarter code to discover the > > binaries. The problem with this approach is that some of these > > tests are just build time. They have no meaning when running on the > > board and need to be filter out somehow. Running the build tests within the Jenkins build makes a lot of sense. Typically, the build tests will have a different command syntax to the runtime tests (otherwise Jenkins would attempt to run both), so filtering should be possible. If the build tests are just a different set of binary blobs from the runtime tests, that may need a fix upstream in Zephyr to distinguish between the two modes. > I see, that makes some sense. But thinking further, I'm not entirely > sure about "build only" tests. Zephyr's sanitycheck test has such > concept, but I'd imagine it comes from the following reasons: a) > sanitycheck runs tests on QEMU, which has very bare hardware support, > so many tests are not runnable; b) sanitycheck can operate on > "samples", not just "tests", as sample can be interactive, etc. it > makes sense to only build them, not run. > > So, I'm not exactly sure about build-only tests on real HW boards. The > "default" idea would be that they should run, but I imagine in > reality, some may need to be filtered out. But then blacklisting > would be better approach than whitelisting. And I'm not sure if > Zephyr has concept of "skipped" tests which may be useful to handle > hardware variations. (Well, I actually dunno if LAVA supports skipped > tests!) Yes, LAVA has support for pass, fail, skip, unknown. For POSIX shell tests, the test writer just calls lava-test-case name --result skip For monitor tests, like Zephyr, it's down to the pattern but skip is as valid as pass and fail (as is unknown) for the result of the matches within the pattern. > Anyway, these are rough ideas for the future. I've spent couple of > weeks of munging with LITE CI setup, there're definitely some > improvements, but also a Pandora box of other ideas and improvements > to make. I'm wrapping up for now, but hope to look again in some time > (definitely hope to look before the Connect, so we can discuss further > steps there). In the meantime, I hope that more boards will be > installed in the Lab and stability of them improves (so far they seem > to be pretty flaky). There are known limitations with the USB subsystem and associated hardware across all architectures, affecting test devices and the workers which run the automation. LAVA has to drive that subsystem very hard for both fastboot devices and IoT devices. There are also problems due to the design of methods like fastboot and some of the IoT support which result from a single-developer model, leading to buggy performance when used at scale and added complexity in deploying workarounds to isolate such protocols in order to prevent interference between tests. The protocols themselves often lack robust error handling or retry support. Other deployment methods which rely on TFTP/network deployments are massively more reliable at scale, so comparing reliability across different device types is problematic. > [] > > > > - test: > > > monitors: > > > - name: foo > > > start: "" > > > end: Hello, ZJS world! > > > pattern: (?P<result>(PASS|FAIL))\s-\s(?P<test_case_id>\w+)\. > > > > > > So, the "start" substring is empty, and perhaps matches a line > > > output by a USB multiplexer or board bootloader. "End" substring > > > is actually the expected single-line output. And "pattern" is > > > unused (dunno if it can be dropped without def file syntax > > > error). Is there a better way to handle single-line test > > > output? > > > > You're making a silent assumption that if there is a matching line, > > the test is passed. In case of other tests (zephyr unit tests), it's > > not the case. The 'start' matches some line which is displayed when > > zephyr is booting. End matches the line which is displayed after all > > testing is done. The pattern follows the unit test pattern. > > Thanks, but I'm not sure I understand this response. I don't challenge > that Zephyr unittests need this support, or the way they're handled. > LITE however needs to test more things than "batch" Zephyr unittests. > I present another usercase which albeit simple, barely supported by > LAVA. (That's a question to LAVA team definitely.) LAVA result handling is ultimately a pattern matching system. Patterns must have a unique and reliable start string and a unique and reliable end string. An empty start string is just going to cause misleading results and bad pattern matches as the reality is that most boards emit some level of random junk immediately upon connection which needs to be ignored. So there needs to be a reliable, unique, start string emitted by the test device. It is not enough to *assume* a start at line zero, doing so increases the problems with reliability. > > > > Well, beyond a simple output matching, it would be nice even for > > > the initial "smoke testing" to actually make some input into the > > > application and check the expected output (e.g., input: "2+2", > > > expected output: "4"). Is this already supported for LAVA "v2" > > > pipeline tests? I may imagine that would be the same kind of > > > functionality required to test bootloaders like U-boot for Linux > > > boards. > > > > I didn't use anything like this in v2 so far, but you're probably > > best off doing sth like > > > > test 2+2=4 PASS. > > > > than you can easily create pattern that will filter the output. In > > case of zephyr pattern is the only way to filter things out as there > > is no shell (?) on the board. > > So, the problem, for starters, is how to make LAVA *feed* the > input, as specified in the test definition (like "2+2") into a board. That will need code changes, so please make a formal request for this support at CTT https://projects.linaro.org/servicedesk/customer/portal/1 so that we can track exactly what is required. > > As there were no reply from LAVA team (I may imagine they're busy with > other things), I decided to create a user story in Jira for them, as I > couldn't create a LAVA-* ticket, I created it as > https://projects.linaro.org/browse/LITE-175 . Hopefully that won't go > unnoticed and LAVA team would get to it eventually. That JIRA story is in the LITE project. Nobody in the LAVA team can manage those stories. It needs a CTT issue which can then be linked to the LITE story and from which a LAVA story can also be linked. Sadly, any story in the LITE project would go completely unnoticed by the LAVA software team until it is linked to CTT so that the work can be prioritised and the relevant LAVA story created. That's just how JIRA works. > > > > > milosz > > Thanks! > > -- > Best Regards, > Paul > > Linaro.org | Open source software for ARM SoCs > Follow Linaro: http://www.facebook.com/pages/Linaro > http://twitter.com/#!/linaroorg - http://www.linaro.org/linaro-blog > _______________________________________________ > linaro-validation mailing list > linaro-validation(a)lists.linaro.org > https://lists.linaro.org/mailman/listinfo/linaro-validation -- Neil Williams ============= http://www.linux.codehelp.co.uk/

8 years, 2 months

2
1
0 0

Re: [Lava-users] Fwd: OE/Yocto testcase

by Robert Berger

Hi, On 2017-07-04 15:18, Agustin Benito Bethencourt wrote: > Dear CIP friends, > > please check below. This is a use case we will meet in a few > weeks/months. It is important to see others walking the same route. My simple tests are running now thanks to the help of the nice people from the #linaro-lava chat. 1) Crazy me decided to use upstream u-boot 2017.05 instead of running some ancient version from 2014;) 1.1) which happens to have a different AUTOBOOT_PROMPT than the one Lava expects "Press SPACE to abort autoboot in %d seconds\n" instead of "Hit any key to stop autoboot" so (since) I would like to be able to stay as close as possible to upstream LAVA 2017.06 I patched u-boot[1]. Note that this could be fixed with LAVA as well - interrupt_prompt: {{ interrupt_prompt|default('Hit any key to stop autoboot') }} 1.2) also the SYS_PROMPT of upstream u-boot is different than the one expected by LAVA and again I made a u-boot patch[2]. Note that this could be fixed with LAVA as well - {% set bootloader_prompt = bootloader_prompt|default('U-Boot') %} 2) After some searching it turned out that LAVA set some funny variables in u-boot which made my kernel crash. (Crazy me decided to use a 4.9.x uImage without baked in address). Adding this: {% set base_high_limits = false %} to my bbb03.jinja2 file fixed it. ... obviously ;) Regards, Robert [1] https://github.com/RobertBerger/meta-mainline/blob/pyro-training-v4.9.x/u-b… [2] https://github.com/RobertBerger/meta-mainline/blob/pyro-training-v4.9.x/u-b…

8 years, 2 months

1
0
0 0

Re: [Lava-users] [Linaro-validation] Further steps with Zephyr tests

by Paul Sokolovsky

Hello Milosz, Thanks for routing this thread to lava-users - when I made initial post to linaro-validation, I check my archive and so that e.g. Neil posts there frequently, but I missed that it's not official LAVA list. On Mon, 3 Jul 2017 22:25:31 +0100 Milosz Wasilewski <milosz.wasilewski(a)linaro.org> wrote: [] > > So, I'm not exactly sure about build-only tests on real HW boards. > > The "default" idea would be that they should run, but I imagine in > > reality, some may need to be filtered out. But then blacklisting > > would be better approach than whitelisting. And I'm not sure if > > Zephyr has concept of "skipped" tests which may be useful to handle > > hardware variations. (Well, I actually dunno if LAVA supports > > skipped tests!) > > As far as I can tell they acutely run on the board, but usually output > just 'Hello world!' or sth similar. As we discussed with Kumar, this > is still OK. What Kumar requested (and I still didn't deliver) is that > whenever the LAVA test job completes, the test should be considered > 'passed'. So we wouldn't have to do any parsing of patterns. I'm not > sure if that will work, but it's worth to try. Hmm, I wonder what would be criteria for being "failed" for such tests... Anyway, thanks for sharing - I'm not familiar with all Zephyr tests/samples myself, will keep in mind such issues when looking into them. [] > > more boards will be installed in the Lab and stability of them > > improves (so far they seem to be pretty flaky). > > > > You're absolutely right. This is a pretty big task to work on and IMHO > requires someone to work full time at least for couple of weeks. The > second part is also true, the boards don't behave as they should. I > guess Dave can elaborate more on that. I can only see the result - > boards (frdm-kw41z) don't run the tests they're requested. Matt Hart actually showed me a ticket on that, so at least it's confirmed/known issue in works. But even with arduino_101 and frdm_k64f, I hit cases more than once when board(s) were stuck for extended time, but still were routed jobs to (which either failed or timed out). So, there may be problem with health checks, which either don't run frequently enough or aren't robust enough. arduino_101 is all the lone one, so if something happens to it, there's no backup. Etc, etc. [] > > So, the problem, for starters, is how to make LAVA *feed* the > > input, as specified in the test definition (like "2+2") into a > > board. > > Right. What I proposed was coding all the inputs in the test itself. Well, that would require bunch of legwork, but the biggest problem is that it wouldn't test what's actually required. E.g. both JerryScript and MicroPython Zephyr ports are actually interactive apps working over serial connection. And functional testing of them would be feeding something over this serial connection and checking that results are as expected. I'll keep in mind idea of "builtin" tests though. Thanks! -- Best Regards, Paul Linaro.org | Open source software for ARM SoCs Follow Linaro: http://www.facebook.com/pages/Linaro http://twitter.com/#!/linaroorg - http://www.linaro.org/linaro-blog

8 years, 2 months

1
0
0 0

Re: [Lava-users] [Linaro-validation] Further steps with Zephyr tests

by Milosz Wasilewski

I was too quick to hit reply. CCing lava-users for comments from LAVA team. milosz On 3 July 2017 at 21:50, Paul Sokolovsky <paul.sokolovsky(a)linaro.org> wrote: > Hello Milosz, > > I appreciate getting at least some response ;-). Some questions however > could use a reply from LAVA team, I guess. > > On Mon, 3 Jul 2017 13:34:49 +0100 > Milosz Wasilewski <milosz.wasilewski(a)linaro.org> wrote: > > [] > >> > jobs submit a number of tests to LAVA (via >> > https://qa-reports.linaro.org/) for the following boards: >> > arduino_101, frdm_k64f, frdm_kw41z, qemu_cortex_m3. Here's an >> > example of cumulative test report for these platforms: >> > https://qa-reports.linaro.org/lite/zephyr-upstream/tests/ >> > >> > That's really great! (Though the list of tests to run in LAVA seems >> > to be hardcoded: >> > https://git.linaro.org/ci/job/configs.git/tree/zephyr-upstream/submit_for_t…) >> > >> >> It is, as I wasn't really sure what to test. The build job needs to >> prepare the test templates to be submitted to LAVA. In case of zephyr >> each tests is a separate binary. So we end up with the number of file >> paths to substitute in the template. Hardcoding was the easiest thing >> to get things running. But I see no reason why it wouldn't be changed >> with some smarter code to discover the binaries. The problem with this >> approach is that some of these tests are just build time. They have no >> meaning when running on the board and need to be filter out somehow. > > I see, that makes some sense. But thinking further, I'm not entirely > sure about "build only" tests. Zephyr's sanitycheck test has such > concept, but I'd imagine it comes from the following reasons: a) > sanitycheck runs tests on QEMU, which has very bare hardware support, > so many tests are not runnable; b) sanitycheck can operate on "samples", > not just "tests", as sample can be interactive, etc. it makes sense to > only build them, not run. > > So, I'm not exactly sure about build-only tests on real HW boards. The > "default" idea would be that they should run, but I imagine in reality, > some may need to be filtered out. But then blacklisting would be better > approach than whitelisting. And I'm not sure if Zephyr has concept of > "skipped" tests which may be useful to handle hardware variations. > (Well, I actually dunno if LAVA supports skipped tests!) > > Anyway, these are rough ideas for the future. I've spent couple of > weeks of munging with LITE CI setup, there're definitely some > improvements, but also a Pandora box of other ideas and improvements to > make. I'm wrapping up for now, but hope to look again in some time > (definitely hope to look before the Connect, so we can discuss further > steps there). In the meantime, I hope that more boards will be > installed in the Lab and stability of them improves (so far they seem > to be pretty flaky). > > [] > >> > - test: >> > monitors: >> > - name: foo >> > start: "" >> > end: Hello, ZJS world! >> > pattern: (?P<result>(PASS|FAIL))\s-\s(?P<test_case_id>\w+)\. >> > >> > So, the "start" substring is empty, and perhaps matches a line >> > output by a USB multiplexer or board bootloader. "End" substring is >> > actually the expected single-line output. And "pattern" is unused >> > (dunno if it can be dropped without def file syntax error). Is >> > there a better way to handle single-line test output? >> >> You're making a silent assumption that if there is a matching line, >> the test is passed. In case of other tests (zephyr unit tests), it's >> not the case. The 'start' matches some line which is displayed when >> zephyr is booting. End matches the line which is displayed after all >> testing is done. The pattern follows the unit test pattern. > > Thanks, but I'm not sure I understand this response. I don't challenge > that Zephyr unittests need this support, or the way they're handled. > LITE however needs to test more things than "batch" Zephyr unittests. I > present another usercase which albeit simple, barely supported by LAVA. > (That's a question to LAVA team definitely.) > >> > Well, beyond a simple output matching, it would be nice even for the >> > initial "smoke testing" to actually make some input into the >> > application and check the expected output (e.g., input: "2+2", >> > expected output: "4"). Is this already supported for LAVA "v2" >> > pipeline tests? I may imagine that would be the same kind of >> > functionality required to test bootloaders like U-boot for Linux >> > boards. >> >> I didn't use anything like this in v2 so far, but you're probably best >> off doing sth like >> >> test 2+2=4 PASS. >> >> than you can easily create pattern that will filter the output. In >> case of zephyr pattern is the only way to filter things out as there >> is no shell (?) on the board. > > So, the problem, for starters, is how to make LAVA *feed* the > input, as specified in the test definition (like "2+2") into a board. > > As there were no reply from LAVA team (I may imagine they're busy with > other things), I decided to create a user story in Jira for them, as I > couldn't create a LAVA-* ticket, I created it as > https://projects.linaro.org/browse/LITE-175 . Hopefully that won't go > unnoticed and LAVA team would get to it eventually. > >> >> milosz > > Thanks! > > -- > Best Regards, > Paul > > Linaro.org | Open source software for ARM SoCs > Follow Linaro: http://www.facebook.com/pages/Linaro > http://twitter.com/#!/linaroorg - http://www.linaro.org/linaro-blog

8 years, 2 months

1
0
0 0

OE/Yocto testcase

by Robert Berger

Hi, I succeeded in running the example Lava tests[1][2] on a Beagle Bone Black in my lab. Now I would like to test a core-image-minimal I built with Yocto and cooked up something like this[3][4]. The result is unfortunately this[5]. I guess one of the problems I have is this: lava-test: # ls -l /lava-38/ export SHELL=/bin/sh /lava-38/bin/lava-test-runner /lava-38/0ls: cannot access '/lava-38/': No such file or directory lava-test: # export SHELL=/bin/sh lava-test: # /lava-38/bin/lava-test-runner /lava-38/0 -sh: /lava-38/bin/lava-test-runner: No such file or directory Regards, Robert [1] https://validation.linaro.org/static/docs/v2/examples/test-jobs/standard-ar… [2] https://validation.linaro.org/static/docs/v2/examples/test-jobs/standard-ar… [3] https://github.com/RobertBerger/git.linaro.org-lava-team-refactoring/blob/m… [4] https://github.com/RobertBerger/git.linaro.org-qa-test-definitions/blob/mas… [5] https://pastebin.com/14pjxhiN

8 years, 2 months

1
0
0 0

job yaml

by Li, Xuelian

Hi,all: If devive have boot successed，can I skip deploy and boot step in job.yaml? This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years, 2 months

3
2
0 0

Is it possible to get data in device dictionary inside a job?

by Li, Yifan2

Hello LAVA experts, I am currently working with LAVA and was asked to find is there a way to get data from device dictionary inside a running job. Details like below: I have a device name demo-01, and its device-dictionary have one line like "{% set my_property = 'my_prop' %}". Then I have a job running on demo-01 device, and I would like to use string 'my_prop' to passed into a script during running. Is it possible to get device-dictionary data directly from job definition(Job Submitter webpage) or test definition(yaml file)? If yes, how could I do this? If not, is there any good way you would like to share to solve this problem? Thanks and Best Regards, Yifan

8 years, 2 months

2
6
0 0

About Test Duration

by kingboli

Hi all, Is there any limitation on test duration when using LAVA to do tests? Recently, I found my tests were automatically canceled after running for 24 hours and I could not find any clue about why the job was canceled. Can anybody give me some help? (Very sorry my LAVA version is V1.) 12794.0 <LAVA_DISPATCHER>2017-06-25 02:31:42 PM INFO: Cancel operation 12794.1 <LAVA_DISPATCHER>2017-06-25 02:31:42 PM DEBUG: finally status fail 12794.2 <LAVA_DISPATCHER>2017-06-25 02:31:42 PM WARNING: [ACTION-E] boot_image is finished with error (Cancel). 12794.3 <LAVA_DISPATCHER>2017-06-25 02:31:42 PM WARNING: Target : power_off being call Thanks.

8 years, 2 months

2
1
0 0

libzmq issue between Debian Jessie and Stretch

by Guillaume Tucker

Hi, We've hit an issue when running lava-master on Debian Jessie and lava-slave on Debian Stretch, after a few minutes the slave would stop working. After some investigation, it turned out to be due to a difference of the libzmq versions in Jessie (4.0.5+dfsg-2) and Stretch (4.2.1-4) causing some protocol errors. The line that detects the error in Stretch is: https://github.com/zeromq/libzmq/blob/7005f22726d4a6ca527f27560a0a132394fdb… This appears to be due to how the "once" counter gets written into memory and into the zmq packets: the libzmq version from Jessie uses memcpy whereas the one in Stretch calls put_uint64. As a result the byte endianness has changed from little to big, causing the packets to work until "once" reaches 255 which translates into 0xff << 56, after which it overflows to 0 and causes the error. This is not a LAVA bug as such, rather a libzmq one, but it impacts interoperability between Jessie and Stretch for LAVA so it may need to be documented or resolved somehow. We've installed the new version of libzmq onto our Jessie servers to align them with Stretch; doing this does fix the problem. Best wishes, Guillaume

8 years, 3 months

2
1
0 0

Lab tool for DUT manual usage,

by Li, Yongtao

Hi All, We have a Lab and the DUTs in the Lab will be used for both automation testing(LAVA) and manual usage(development, debug, manual testing). And we will develop a tool for the manual usage, please find the basic requirements of the tool in the attachment. I list possible solutions about how to develop the lab tool and let the tool cooperate with LAVA on the same Lab hardwares. which one is better? could you give your suggestions? 1. Develop the lab tool based on LAVA framawork. Need to modify LAVA DB: add some new tables for network ports, port attribution, port connection, usage log, notes, also need to add columns to existing tables lava_scheduler_app_device. Need to modify LAVA command line, add functions like lava-server reserve dut1, lava-server connect dut1. Need to add new codes for the features which LAVA doesn't support, part of codes may be reused, I need to look into LAVA and check how to reuse them. The tool will be developed based on https://github.com/Linaro/lava-server and will be installed on LAVA master, right? Most probably we will maintain the codes change in local repository because it is difficult to upstream thousands of codes changes to linaro master repository. We need to merge the changes in master repository to local repository. 2. Develop the lab tool as a separated tool and use separated DB. There are two ways to avoid the DUT conflicts between the lab tool and LAVA: a) Lab tool DB maintain all DUTs, LAVA DB also maintain all DUTs, when a user want to do automation testing on one DUT, he need to reserve this DUT via lab tool in advance, then he can run automation testing on the specified DUT. b),Divide the DUTs to two groups, one group is for automation testing and it will be added to LAVA DB, the other one group is for manual usage and will be added to lab tool DB. Another question about how to link/unlink two network ports dynamically(see requirement#7 in attachment) in automation testing. I am not sure whether LAVA can support this feature later, the simple way is supporting it in test script: Do the link/unlink in test script: subprocess.Popen(["lab-tool","link","dut1.eth_1g_1","dut2.eth_1g_1"]) , and test script needs to get DUT name from LAVA in advance. Does this work? BR Yongtao This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years, 3 months

2
2
0 0

LAVA v2 multinode test with hikey

by Arthur She

Hi, I am trying to run a multinode test job <https://staging.validation.linaro.org/scheduler/job/175378/definition> in LAVA v2. It failed <https://staging.validation.linaro.org/scheduler/job/175378#L555>. After a short discussion with Chase, I realized if I want to run multinode test job I have to use "lava-multinode" protocol. However, I still need "lava-lxc" to deploy the test image for hikey. I am wondering if it is possible to use two protocols, lava-lxc for deployment test image and lava-multinode to run the test, in the same test job? If yes, could you please provide some examples? Thanks, Arthur

8 years, 3 months

4
16
0 0

if i can add the device_tags by the cmdline ?

by Xu, Hongyu

Hi , I known can add the device_tags by the web , and if i can add the device_tags by the cmdline ? Best Regards XuHongyu This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years, 3 months

2
1
0 0

deploy a x86 device

by Li, Xuelian

Hello，Neil： 1.Can you give me a deploy flow for a x86 device . 2．I want to use web UI submit job for one time N jobs. (N>10) How can I do it? Thanks. Amanda This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years, 3 months

3
7
0 0

Can be used multi database in the lava ?

by Xu, Hongyu

Hi all, Can be used multi database in the lava ? And is there any introduction to this ? Best Regards XuHongyu This email is intended only for the named addressee. It may contain information that is confidential/private, legally privileged, or copyright-protected, and you should handle it accordingly. If you are not the intended recipient, you do not have legal rights to retain, copy, or distribute this email or its contents, and should promptly delete the email and all electronic copies in your system; do not retain copies in any media. If you have received this email in error, please notify the sender promptly. Thank you.

8 years, 3 months

2
1
0 0

LAVA installation in jessie and jessie-backports

by Senthil Kumaran S

Hi, Someone with the nickname "Nurav" pinged in #linaro-lava IRC channel last week regarding installation problem to install lava in jessie using the jessie-backports repository. The person also diligently followed up with me to check what is wrong with his/her installation on private messages. I did find sometime to do the testing today. Since, I do not know any contact details of 'Nurav' I am writing my findings here, based on lava installation I did on fresh jessie containers, hopefully the person is available in this mailing list and will see this message. The fresh installation went fine with both jessie and also using jessie backports and LAVA production-repo. I ve put up the complete log of the installations I tried in the following files: * Fresh installation of lava from jessie-backports - http://people.linaro.org/~senthil.kumaran/logs/lava-jessie-bp-installation.… * Installing 2017.5 lava from jessie to jessie-backports to production-repo - http://people.linaro.org/~senthil.kumaran/logs/lava-jessie-bp-installation-… HTH. Thank You. -- Senthil Kumaran http://www.stylesen.org/ http://www.sasenthilkumaran.com/

8 years, 3 months

1
0
0 0

[LAVA V2] how to submit job to one specified device

by wangbo

In LAVA V1, we can use `target: {device hostname}` to submit job to one specified device, but I find V2 does not support this. ------------------ Best Regards. Bo Wang

8 years, 3 months

2
1
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

lava-users