From: Robert Hu <robert.hu@vmm.sh.intel.com>
To: Ian Campbell <ian.campbell@citrix.com>
Cc: wei.liu2@citrix.com, "longtao.pang" <longtaox.pang@intel.com>,
robert.hu@intel.com, Ian.Jackson@eu.citrix.com,
xen-devel@lists.xen.org
Subject: Re: [OSSTEST Nested PATCH v8 4/7] Changes on test step of Debian hvm guest install
Date: Thu, 23 Apr 2015 13:59:21 +0800 [thread overview]
Message-ID: <1429768761.31307.17.camel@localhost> (raw)
In-Reply-To: <1429612114.4743.29.camel@citrix.com>
On Tue, 2015-04-21 at 11:28 +0100, Ian Campbell wrote:
> On Mon, 2015-04-13 at 17:19 -0400, longtao.pang wrote:
> > 1. Increase disk size to accommodate to nested test requirement.
> > 2. Since 'Debain-xxx-.iso' image will be stored in rootfs of L1 guest,
> > therefore needs more disk capacity, increase root partition size in
> > preseed generation.
> > 3. In L1 installation context, assign more memory to it; Since it
> > acts as a nested hypervisor anyway.
> > 4. Comment out CDROM entry in sources.list to make HTTP URL entry
> > available for L1 hvm guest.
> > 5. Enable nestedhvm feature in ExtraConfig for nested job.
> >
> > Signed-off-by: longtao.pang <longtaox.pang@intel.com>
> > ---
> > Changes in v8:
> > 1. Update a conventional way to comment out CDROM entry in sources.list.
> > 2. Enable nestedhvm feature only for the nested job.
> > 3. Set nested disk and memroy size to be driven from runvars in
> > 'ts-debian-hvm-install'.
> > ---
> > ts-debian-hvm-install | 9 ++++++---
> > 1 file changed, 6 insertions(+), 3 deletions(-)
> >
> > diff --git a/ts-debian-hvm-install b/ts-debian-hvm-install
> > index cfd5144..13009d0 100755
> > --- a/ts-debian-hvm-install
> > +++ b/ts-debian-hvm-install
> > @@ -36,7 +36,7 @@ our $ho= selecthost($whhost);
> >
> > # guest memory size will be set based on host free memory, see below
> > our $ram_mb;
> > -our $disk_mb= 10000;
> > +our $disk_mb= $r{'nested_disk'} ? $r{'nested_disk'} : 10000;
>
> This shouldn't hardcode 'nested' here, but use the guest ident. (nested
> is the wrong name now anyway I think, since it is nestedl1 in this
> iteration, which is why we avoid such hardcoding)
Right, we shall store this runvar with name of ${l1_ident}_disksize in
ts-nested-setup.
>
> So you should arrange to do this somewhere that $gho is in scope and use
> guest_var($gho, 'disk', 10000). $gho is in scope in $prep where $disk_mb
> is used and AFAICT it is only used in that function, so I think you can
> move this to a local variable
yes, it is supposed to be local of prep(). However, look at other ts-*
(ts-debian-install, ts-freebsd-install, etc.), $disk_mb is defined
'our'. They seems to be in alliance, are we going to it specifically in
ts-debian-hvm-install? or in some future day, we change these
ts-*-install in another patch series?
To use guest_var(), it needs $gho in scope as you said. But in prep() we
can see until prepareguest() returns, we don't have $gho in scope, while
prepareguest() needs $disk_mb as input. So here we get into a loop.
I think we shall in prepareguest(), check if $r{${guest_ident}_disksize}
is defined, then override passed in $mb value. How would you like this?
>
> Perhaps also consider a more descriptive runvar name too, like disksize?
>
> > our $guesthost= "$gn.guest.osstest";
> > our $gho;
> > @@ -60,7 +60,7 @@ d-i partman-auto/expert_recipe string \\
> > use_filesystem{ } filesystem{ vfat } \\
> > mountpoint{ /boot/efi } \\
> > . \\
> > - 5000 50 5000 ext4 \\
> > + 10000 50 10000 ext4 \\
> > method{ format } format{ } \\
> > use_filesystem{ } filesystem{ ext4 } \\
> > mountpoint{ / } \\
> > @@ -75,6 +75,7 @@ d-i preseed/late_command string \\
> > in-target mkdir -p /boot/efi/EFI/boot; \\
> > in-target cp /boot/efi/EFI/debian/grubx64.efi /boot/efi/EFI/boot/bootx64.efi ;\\
> > in-target mkdir -p /root/.ssh; \\
> > + in-target sed -i 's/^deb *cdrom/#&/g' /etc/apt/sources.list; \\
> > in-target sh -c "echo -e '$authkeys'> /root/.ssh/authorized_keys";
> > END
> > return $preseed_file;
> > @@ -149,9 +150,11 @@ sub prep () {
> > target_putfilecontents_root_stash($ho, 10, preseed(),
> > $preseed_file_path);
> >
> > + my $nestedhvm_feature = $gn eq 'nestedl1' ? 'nestedhvm=1' : '';
>
> This should be based on guest_var($gho, "enable_nestedhvm", undef) not
> $gn eq "nested1".
>
> Also I think this would be better done as:
> my $extra_config = '';
> $extra_config .= 'nestedhvm=1\n'
> if guest_var($gho, "enable_nestedhvm", 'false') eq 'true';
>
> Then use $extra_config below. This will make it easier to add more such
> things in the future.
Ack.
>
> > more_prepareguest_hvm($ho,$gho, $ram_mb, $disk_mb,
> > OnReboot => 'preserve',
> > Bios => $r{bios},
> > + ExtraConfig => $nestedhvm_feature,
> > PostImageHook => sub {
> > my $cmds = iso_copy_content_from_image($gho, $newiso);
> > $cmds .= prepare_initrd($initrddir,$newiso,$preseed_file_path);
> > @@ -174,7 +177,7 @@ my $ram_lots = 5000;
> > if ($host_freemem_mb > $ram_lots * 2 + $ram_minslop) {
> > $ram_mb = $ram_lots;
> > } else {
> > - $ram_mb = 768;
> > + $ram_mb = $r{"${gn}_memory"} ? $r{"${gn}_memory"} : 768;
>
> guest_var again and I think this can be local to prep() too..
Also, agree to use guest_var() while may be still be global here as
other similar ts-*-install does, change them together in the future.
>
> > }
> > logm("Host has $host_freemem_mb MB free memory, setting guest memory size to $ram_mb MB");
> >
>
>
next prev parent reply other threads:[~2015-04-23 5:59 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-13 21:19 [OSSTEST Nested PATCH v8 0/7] Introduction of netsted HVM test job longtao.pang
2015-04-13 21:19 ` [OSSTEST Nested PATCH v8 1/7] parsing grub which has 'submenu' primitive longtao.pang
2015-04-21 10:12 ` Ian Campbell
2015-04-13 21:19 ` [OSSTEST Nested PATCH v8 2/7] Changes to support '/boot' leading paths of kernel, xen, in grub longtao.pang
2015-04-21 10:13 ` Ian Campbell
2015-04-13 21:19 ` [OSSTEST Nested PATCH v8 3/7] Edit some APIs in TestSupport.pm for nested test longtao.pang
2015-04-21 10:19 ` Ian Campbell
2015-04-21 12:33 ` Ian Jackson
2015-04-21 12:53 ` Ian Campbell
2015-04-21 13:28 ` Ian Jackson
2015-04-21 13:41 ` Ian Campbell
2015-04-21 14:30 ` Ian Jackson
2015-04-21 14:43 ` Ian Campbell
2015-04-22 8:25 ` Pang, LongtaoX
2015-04-22 9:48 ` Ian Campbell
2015-04-22 12:50 ` Ian Jackson
2015-04-23 0:34 ` Hu, Robert
2015-04-27 9:36 ` Robert Hu
2015-04-28 7:41 ` Ian Campbell
2015-04-13 21:19 ` [OSSTEST Nested PATCH v8 4/7] Changes on test step of Debian hvm guest install longtao.pang
2015-04-21 10:28 ` Ian Campbell
2015-04-23 5:59 ` Robert Hu [this message]
2015-04-23 6:52 ` Ian Campbell
2015-04-23 10:43 ` Hu, Robert
2015-04-23 12:04 ` Ian Campbell
2015-04-23 11:07 ` Ian Jackson
2015-04-13 21:19 ` [OSSTEST Nested PATCH v8 5/7] Add new script to customize nested test configuration longtao.pang
2015-04-21 10:40 ` Ian Campbell
2015-04-22 8:35 ` Pang, LongtaoX
2015-04-22 9:56 ` Ian Campbell
2015-04-23 9:38 ` Robert Hu
2015-04-23 11:30 ` Ian Campbell
2015-04-23 13:05 ` Ian Campbell
2015-04-24 8:45 ` Pang, LongtaoX
2015-04-28 7:39 ` Ian Campbell
2015-04-23 7:27 ` Pang, LongtaoX
2015-04-23 11:35 ` Ian Campbell
2015-04-13 21:19 ` [OSSTEST Nested PATCH v8 6/7] Compose the main recipe of nested test job longtao.pang
2015-04-21 10:48 ` Ian Campbell
2015-04-22 8:38 ` Pang, LongtaoX
2015-04-22 11:04 ` Ian Campbell
2015-04-22 11:23 ` Ian Campbell
2015-04-23 8:08 ` Pang, LongtaoX
2015-04-23 11:05 ` Ian Jackson
2015-04-24 6:31 ` Robert Hu
2015-04-23 10:49 ` Robert Hu
2015-04-13 21:19 ` [OSSTEST Nested PATCH v8 7/7] Add test job for nest test case longtao.pang
2015-04-21 10:51 ` Ian Campbell
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1429768761.31307.17.camel@localhost \
--to=robert.hu@vmm.sh.intel.com \
--cc=Ian.Jackson@eu.citrix.com \
--cc=ian.campbell@citrix.com \
--cc=longtaox.pang@intel.com \
--cc=robert.hu@intel.com \
--cc=wei.liu2@citrix.com \
--cc=xen-devel@lists.xen.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.