Hi there. We'd like to run a Fast Model in the validation lab for KVM testing. Is there a blueprint for this? What's the status?
Paul and I discussed a rough plan a few months ago. It was along the lines of: * A x86 machine as the Fast Model host * An emulated vexpress-a15 as the KVM host * A vexpress-a15 as the KVM guest * LAVA treats the Fast Model as a board * Jobs are spawned into the LAVA scheduler * Once the KVM host is running, everything else is toolchain specific and done via shell scripts
The dispatcher would: * Grab the hwpack * Grab the nano rootfs * Build the rootfs with separate kernel, initrd, and dtb using linaro-media-create * Start the Fast Model with the boot wrapper, kernel, and rootfs * Use the console to run the test script
There's more information on the steps required at: https://wiki.linaro.org/MichaelHope/Sandbox/KVMUseCase
-- Michael
Hi.
We'll see some beta deployments this week. I have not tried doing any KVM testing. What I can currently offer you is vexpress x4 or x1 (both A7+A15) that you can run lava-test's on. If you want we can also add a simple "run this shell" action for experiments. Note: keep in mind that it's not "ready" yet, it's not the same dispatcher code you can use on physical boards and it has some different properties. You may be interested in https://blueprints.launchpad.net/lava-core/+spec/dispatcher-transition-plan
Thanks ZK
On Mon, May 7, 2012 at 5:40 AM, Michael Hope michael.hope@linaro.org wrote:
Hi there. We'd like to run a Fast Model in the validation lab for KVM testing. Is there a blueprint for this? What's the status?
Paul and I discussed a rough plan a few months ago. It was along the lines of: * A x86 machine as the Fast Model host * An emulated vexpress-a15 as the KVM host * A vexpress-a15 as the KVM guest * LAVA treats the Fast Model as a board * Jobs are spawned into the LAVA scheduler * Once the KVM host is running, everything else is toolchain specific and done via shell scripts
The dispatcher would: * Grab the hwpack * Grab the nano rootfs * Build the rootfs with separate kernel, initrd, and dtb using linaro-media-create * Start the Fast Model with the boot wrapper, kernel, and rootfs * Use the console to run the test script
There's more information on the steps required at: https://wiki.linaro.org/MichaelHope/Sandbox/KVMUseCase
-- Michael
linaro-validation mailing list linaro-validation@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-validation
On 7 May 2012 20:10, Zygmunt Krynicki zygmunt.krynicki@linaro.org wrote:
Hi.
We'll see some beta deployments this week. I have not tried doing any KVM testing. What I can currently offer you is vexpress x4 or x1 (both A7+A15) that you can run lava-test's on. If you want we can also add a simple "run this shell" action for experiments. Note: keep in mind that it's not "ready" yet, it's not the same dispatcher code you can use on physical boards and it has some different properties. You may be interested in https://blueprints.launchpad.net/lava-core/+spec/dispatcher-transition-plan
OK. Do you have documentation or a specification you're working to? Does your plan cover my needs or something close to it?
Thanks ZK
On Mon, May 7, 2012 at 5:40 AM, Michael Hope michael.hope@linaro.org wrote:
Hi there. We'd like to run a Fast Model in the validation lab for KVM testing. Is there a blueprint for this? What's the status?
Paul and I discussed a rough plan a few months ago. It was along the lines of: * A x86 machine as the Fast Model host * An emulated vexpress-a15 as the KVM host * A vexpress-a15 as the KVM guest * LAVA treats the Fast Model as a board * Jobs are spawned into the LAVA scheduler * Once the KVM host is running, everything else is toolchain specific and done via shell scripts
The dispatcher would: * Grab the hwpack * Grab the nano rootfs * Build the rootfs with separate kernel, initrd, and dtb using linaro-media-create * Start the Fast Model with the boot wrapper, kernel, and rootfs * Use the console to run the test script
There's more information on the steps required at: https://wiki.linaro.org/MichaelHope/Sandbox/KVMUseCase
On Tue, 8 May 2012 15:56:27 +1200, Michael Hope michael.hope@linaro.org wrote:
On 7 May 2012 20:10, Zygmunt Krynicki zygmunt.krynicki@linaro.org wrote:
Hi.
We'll see some beta deployments this week. I have not tried doing any KVM testing. What I can currently offer you is vexpress x4 or x1 (both A7+A15) that you can run lava-test's on. If you want we can also add a simple "run this shell" action for experiments. Note: keep in mind that it's not "ready" yet, it's not the same dispatcher code you can use on physical boards and it has some different properties. You may be interested in https://blueprints.launchpad.net/lava-core/+spec/dispatcher-transition-plan
OK. Do you have documentation or a specification you're working to?
Not really :/
Does your plan cover my needs or something close to it?
As I understand your plans, yes, you're covered. "Use the console to run the test script" can be done in a large variety of ways of course and I don't know what will best meet your needs there. How would you like the results to be presented?
Cheers, mwh
On 9 May 2012 12:14, Michael Hudson-Doyle michael.hudson@canonical.com wrote:
On Tue, 8 May 2012 15:56:27 +1200, Michael Hope michael.hope@linaro.org wrote:
On 7 May 2012 20:10, Zygmunt Krynicki zygmunt.krynicki@linaro.org wrote:
Hi.
We'll see some beta deployments this week. I have not tried doing any KVM testing. What I can currently offer you is vexpress x4 or x1 (both A7+A15) that you can run lava-test's on. If you want we can also add a simple "run this shell" action for experiments. Note: keep in mind that it's not "ready" yet, it's not the same dispatcher code you can use on physical boards and it has some different properties. You may be interested in https://blueprints.launchpad.net/lava-core/+spec/dispatcher-transition-plan
OK. Do you have documentation or a specification you're working to?
Not really :/
Does your plan cover my needs or something close to it?
As I understand your plans, yes, you're covered. "Use the console to run the test script" can be done in a large variety of ways of course and I don't know what will best meet your needs there. How would you like the results to be presented?
No idea :) The important bit is that we hand off to a shell script (or some other test runner) that takes care of the guest and any other tricky things. The minimum is to gather the output and report a global pass/fail based on the return code.
I'm happy to use whatever you already have defined,
-- Michael
Michael Hope michael.hope@linaro.org writes:
On 9 May 2012 12:14, Michael Hudson-Doyle michael.hudson@canonical.com wrote:
On Tue, 8 May 2012 15:56:27 +1200, Michael Hope michael.hope@linaro.org wrote:
On 7 May 2012 20:10, Zygmunt Krynicki zygmunt.krynicki@linaro.org wrote:
Hi.
We'll see some beta deployments this week. I have not tried doing any KVM testing. What I can currently offer you is vexpress x4 or x1 (both A7+A15) that you can run lava-test's on. If you want we can also add a simple "run this shell" action for experiments. Note: keep in mind that it's not "ready" yet, it's not the same dispatcher code you can use on physical boards and it has some different properties. You may be interested in https://blueprints.launchpad.net/lava-core/+spec/dispatcher-transition-plan
OK. �Do you have documentation or a specification you're working to?
Not really :/
Does your plan cover my needs or something close to it?
As I understand your plans, yes, you're covered. �"Use the console to run the test script" can be done in a large variety of ways of course and I don't know what will best meet your needs there. �How would you like the results to be presented?
No idea :) The important bit is that we hand off to a shell script (or some other test runner) that takes care of the guest and any other tricky things. The minimum is to gather the output and report a global pass/fail based on the return code.
I'm happy to use whatever you already have defined,
I think wrapping this up in the lava-test framework probably makes sense. It's a bit overkill I guess to just gather the output and create a single test result based on exit status, but I think it would integrate most neatly with everything else.
Cheers, mwh
linaro-toolchain@lists.linaro.org
-
Michael Hope -
Michael Hudson-Doyle -
Zygmunt Krynicki