Merge lp:~saviq/simplestreams/add-vhd-images into lp:simplestreams

[Michał Sawicz]

Include VHD images in the stream

Note this suffers from bug #1686750

[Scott Moser (smoser)]

The patch here looks fine, but I wont merge it without someone on cpc team approval.

I do have the general dislike of Ubuntu publishing .vmdk and .qcow2 and
 .here-is-another-format-that-a-hypervisor-could-convert-from-the-open-standard-qcow2.

But that is my only objection, and based on the many formats we publish, we've decided not to use Ubuntu's leverage to help the industry with interoperability.

[lp:~saviq/simplestreams/add-vhd-images]

450. By Michał Sawicz on 2017-04-27

Include VHD images in the stream

[Server Team CI bot (server-team-bot)]

PASSED: Continuous integration, rev:450
https://jenkins.ubuntu.com/server/job/simplestreams-ci/33/
Executed test runs:
    SUCCESS: https://jenkins.ubuntu.com/server/job/simplestreams-ci/nodes=metal-amd64/33
    SUCCESS: https://jenkins.ubuntu.com/server/job/simplestreams-ci/nodes=metal-arm64/33
    SUCCESS: https://jenkins.ubuntu.com/server/job/simplestreams-ci/nodes=metal-ppc64el/33
    SUCCESS: https://jenkins.ubuntu.com/server/job/simplestreams-ci/nodes=metal-s390x/33
    SUCCESS: https://jenkins.ubuntu.com/server/job/simplestreams-ci/nodes=vm-i386/33

Click here to trigger a rebuild:
https://jenkins.ubuntu.com/server/job/simplestreams-ci/33/rebuild

review: Approve (continuous-integration)

[Scott Moser (smoser)]

Is this still necessary?
Desireable?

review: Needs Information

[Michał Sawicz (saviq)]

I've decided to convert from qcow2 because of the large size (30G) of VHD images, so I'm fine dropping this.

[Scott Moser (smoser)]

Hi.
We've moved simplestreams from bzr on launchpad to git on launchpad.
I'm going to mark this merge proposal as Rejected based only on
the fact that it is a bzr based merge proposal.

Please feel free to re-submit the merge proposal using the launchpad git.
Hopefully the cloud-init doc at
 http://cloudinit.readthedocs.io/en/latest/topics/hacking.html
should give you an idea on how to do that.

Unmerged revisions

450. By Michał Sawicz on 2017-04-27

Include VHD images in the stream

=== modified file 'tools/make-test-data'
--- tools/make-test-data	2016-11-28 14:54:35 +0000
+++ tools/make-test-data	2017-04-27 14:45:47 +0000
@@ -63,6 +63,8 @@
     'squashfs': {'size': 12400},
     'squashfs.manifest': {'size': 12401},
     'img': {'size': 12402},
+    'disk1.vhd.zip': {'size': 12403},
+    'vhd.zip': {'size': 12404},
 }
 
 EC2_ENDPOINTS = {

=== modified file 'tools/toolutil.py'
--- tools/toolutil.py	2016-08-02 20:54:11 +0000
+++ tools/toolutil.py	2017-04-27 14:45:47 +0000
@@ -111,6 +111,20 @@
         if rel == "xenial" and serial <= "20160420.3":
             return False
 
+    if suffix == "-disk1.vhd.zip":
+        # -disk1.vhd.zip only available in xenial
+        if codename_cmp(rel, "==", "xenial"):
+            return False
+        if arch != "amd64":
+            return False
+
+    if suffix == ".vhd.zip":
+        # .vhd.zip started in yakkety
+        if codename_cmp(rel, "<", "yakkety"):
+            return False
+        if arch != "amd64":
+            return False
+
     # if some data in /query is not truely available, fill up this array
     # to skip it. ex: export BROKEN="precise/20121212.1 quantal/20130128.1"
     broken = os.environ.get("BROKEN", "").split(" ")
@@ -129,7 +143,8 @@
 
     suffixes = (".tar.gz", "-root.tar.gz", "-disk1.img", "-uefi1.img",
                 ".manifest", ".ova", "-root.tar.xz", "-lxd.tar.xz",
-                ".squashfs", ".squashfs.manifest", ".img",)
+                ".squashfs", ".squashfs.manifest", ".img", "-disk1.vhd.zip",
+                ".vhd.zip",)
     streams = [f[0:-len(".latest.txt")]
                for f in os.listdir(path) if f.endswith("latest.txt")]