- Devel - oVirt List Archives

Fwd: Lago v0.39 is out!
by Nadav Goldin 04 Jun '17

04 Jun '17

---------- Forwarded message ---------- From: Nadav Goldin <ngoldin(a)redhat.com> Date: Sun, Jun 4, 2017 at 2:24 PM Subject: Lago v0.39 is out! To: lago-devel(a)ovirt.org On behalf of the Lago team, I'm pleased to announce the new release of Lago and Lago-ost-plugin: Lago - v0.39 Lago-ost-plugin v0.41 This is the first release where we've separated Lago and lago-ost-plugin(aka ovirtlago) to different repositories, installation procedures should be the same. However, from now on the lago-ost-plugin will follow a different release cycle. It's repository can be found at [1], and docs at [2]. Note that 'lago-ost-plugin' requires Lago >= 0.39. What's new ========= Lago ------- 1. Improved Ansible inventory support. For more details see [3]. 2. Lago SDK - Allows to run most CLI operations directly from Python. See [4] for the docs, and [5] for an example. This is mostly standardization of the already provided SDK. 3. Debian network support in bootstrap stage. Lago Images ----------------- 3 New images were added, please help in verifying them: 1. fc25-base 2. debian8-base 3. ubuntu16.04-base There is a known issue with host name resolution after boot in debian, but it does not affect connectivity. Tests/CI ------------ 1. Moved to tox to setup the virtualenv during the tests: * `tox -e docs` - builds the docs. * `tox -e py27` - run unittests and linters. * `tox -c tox-sdk.ini -- --stage check_patch/check_merged` - to run the functional SDK tests for each stage(after you have installed lago - either in a nested virtualenv or from RPMs). 2. Added SDK functional tests: * Easy to run sanity check while developing, under tests/functional-sdk run: `pytest -s -vvv --setup-show --stage check_patch test_sdk_sanity.py` 3. Added multi-distro tests, which means prior to merging every Lago patch, we'll ensure the core images in templates.ovirt.org are functional with the new patch. 4. Added ansible functional tests on check-merged. For the full changelog see [6]. Upgrading ======== To upgrade using yum or dnf, simply run: ``` yum/dnf update lago ``` Resources ======== Lago Docs: http://lago.readthedocs.io/en/latest/ GitHub: https://github.com/lago-project/lago/ YUM Repository: http://resources.ovirt.org/repos/lago/stable/0.0/rpm/ OST Docs: http://ovirt-system-tests.readthedocs.io/en/latest/ As always, if you find any problems, please open an issue in the GitHub page. Enjoy! Nadav. [1] https://github.com/lago-project/lago-ost-plugin [2] http://lago-ost-plugin.readthedocs.io/en/latest/ [3] https://github.com/lago-project/lago/pull/544 [4] http://lago.readthedocs.io/en/stable/SDK.html [5] https://github.com/lago-project/lago/blob/master/docs/examples/lago_sdk_one… [6] https://github.com/lago-project/lago/compare/0.38...0.39

1 0

[VDSM] flake8 3.0.4 failing vdsm builds
by Nir Soffer 02 Jun '17

02 Jun '17

Hi all, flake8 started to fail today with import error, failing vdsm builds with this error: Traceback (most recent call last): File "/home/nsoffer/src/vdsm/.tox/flake8/bin/flake8", line 7, in <module> from flake8.main.cli import main File "/home/nsoffer/src/vdsm/.tox/flake8/lib/python2.7/site-packages/flake8/main/cli.py", line 2, in <module> from flake8.main import application File "/home/nsoffer/src/vdsm/.tox/flake8/lib/python2.7/site-packages/flake8/main/application.py", line 14, in <module> from flake8.main import options File "/home/nsoffer/src/vdsm/.tox/flake8/lib/python2.7/site-packages/flake8/main/options.py", line 3, in <module> from flake8.main import vcs File "/home/nsoffer/src/vdsm/.tox/flake8/lib/python2.7/site-packages/flake8/main/vcs.py", line 4, in <module> from flake8.main import mercurial File "/home/nsoffer/src/vdsm/.tox/flake8/lib/python2.7/site-packages/flake8/main/mercurial.py", line 7, in <module> import configparser File "/usr/lib/python2.7/site-packages/configparser.py", line 12, in <module> from backports.configparser import ( ImportError: No module named configparser I posted this patch updating tox to use flake8 3.1.0: https://gerrit.ovirt.org/77727 Please rebase on this patch if your build fail. Thanks, Nir

1 0

Using more specific types in code visible to GWT frontend
by Vojtech Szocs 02 Jun '17

02 Jun '17

Hello, in GWT world, being more specific (e.g. using ArrayList instead of just List) is preferred since the GWT compiler doesn't need to introspect all possible subtypes and include them in the generated JavaScript. Using too "generic" types is one of the main reasons of significant GWT-generated JavaScript size. I'd like to improve this and I'm wondering what others think. For example, in BusinessEntity interface, is there any reason why <T extends Serializable> and not e.g. <T extends Guid> ? If the GWT compiler sees a reference to Serializable type and it isn't obvious what's the instance type behind it, it includes *all* Serializable subtypes into the generated JavaScript. You can probably imagine how many of those are visible to GWT compiler via Java classpath. Another example is query & action result allowing arbitrary List/Map/Set impl. to be passed through GWT RPC. What about limiting us to ArrayList, HashSet and HashMap for GWT RPC transfer? Please be aware that usual Java performance tips do not apply to GWT world due to Java-to-JS compilation. (And therefore, using LinkedList instead of ArrayList has no practical effect aside from bigger JS footprint.) I'd like both frontend and backend maintainers to share their thoughts. Alexander's UI redesign efforts [1] are well on their way into the master branch and it's a good opportunity to improve existing GWT frontend, making it smaller & faster for our users. [1] https://gerrit.ovirt.org/#/c/75669/ Thanks! Vojtech

2 2

Engine query & action result classes are now final
by Vojtech Szocs 02 Jun '17

02 Jun '17

Hi everyone, yesterday, we merged https://gerrit.ovirt.org/#/c/75705/ which seals Engine query & action result classes. We will follow this up with renaming "VdcReturnValueBase" to "ActionReturnValue" and doing similar renames to existing "VdcSomething" classes. The reason for finalizing those result classes is to avoid us implementing a GWT custom field serializer for each such subclass, therefore limiting our exposure to GWT RPC specifics when transferring objects between client and server. On a side note, we also upgraded GWT 2.8.0 to 2.8.1 which fixes a GWT compiler bug we had to work around: https://gerrit.ovirt.org/#/c/77264/ Regards, Vojtech

1 1

Regarding site bug fix
by shubham dubey 31 May '17

31 May '17

Hi, I want to fix some issues in ovirt site. Since I never did a pull request before, hence want to know few things. 1)can I fix multiple bug and send a single pr for them. 2) when I fix a issue did I need to write something in the comment of the bug (like 'fixed' or 'patch send'). thanks in advance and I will appreciate for extra related info:) Shubham

2 3

Wednesday Website
by Sandro Bonazzola 31 May '17

31 May '17

Hi, oVirt website issues are piling up, currently counting 63 issues[1] reported. I'd like to propose to dedicate one hour next Wednesday, picking one of them and trying to address it. Most of them doesn't require programming skills. [1] https://github.com/oVirt/ovirt-site/issues -- SANDRO BONAZZOLA ASSOCIATE MANAGER, SOFTWARE ENGINEERING, EMEA ENG VIRTUALIZATION R&D Red Hat EMEA <https://www.redhat.com/> <https://red.ht/sig> TRIED. TESTED. TRUSTED. <https://redhat.com/trusted>

1 0

ovirt-engine-nodejs failing build on fc26
by Sandro Bonazzola 31 May '17

31 May '17

Hi, just an heads up that ovirt-engine-nodejs fails to build on fc26: http://jenkins.ovirt.org/job/ovirt-engine-nodejs_master_build-artifacts-fc2… -- SANDRO BONAZZOLA ASSOCIATE MANAGER, SOFTWARE ENGINEERING, EMEA ENG VIRTUALIZATION R&D Red Hat EMEA <https://www.redhat.com/> <https://red.ht/sig> TRIED. TESTED. TRUSTED. <https://redhat.com/trusted>

3 4

[VDSM] Fedora rawhide build - works!
by Nir Soffer 31 May '17

31 May '17

Hi all, We have now a Fedora rawhide build on travis: https://travis-ci.org/nirs/vdsm/builds/231649060 To make it work, I had to: - disable kvm2ovirt import test, since ovirt-imageio-common is not available for rawhide on ovirt repos Building ovirt-imageio for rawhide should be easy - disable 2 tests importing blivet, since importing it fail with ValueError on rawhide (blivet bug?) The build is using ovirtorg/vdsm-test-fedora-rawhide image. Please review: https://gerrit.ovirt.org/#/q/topic:rawhide Cheers, Nir

2 3

Re: [ovirt-devel] Making event types backwards compatible?
by Oved Ourfali 30 May '17

30 May '17

On May 30, 2017 11:28, "Juan Hernández" <jhernand(a)redhat.com> wrote: On 05/30/2017 09:38 AM, Tomas Jelinek wrote: > > > On Tue, May 30, 2017 at 9:16 AM, Juan Hernández <jhernand(a)redhat.com > <mailto:jhernand@redhat.com>> wrote: > > On 05/30/2017 08:55 AM, Tomas Jelinek wrote: > > > > > > On Tue, May 30, 2017 at 7:20 AM, Michal Skrivanek < mskrivan(a)redhat.com <mailto:mskrivan@redhat.com> > > <mailto:mskrivan@redhat.com <mailto:mskrivan@redhat.com>>> wrote: > > > > > On 29 May 2017, at 11:44, Juan Hernández <jhernand(a)redhat.com <mailto:jhernand@redhat.com> > <mailto:jhernand@redhat.com <mailto:jhernand@redhat.com>>> wrote: > > > > > >> On 05/29/2017 11:27 AM, Michal Skrivanek wrote: > > >> > > >>> On 29 May 2017, at 10:39, Juan Hernández < jhernand(a)redhat.com <mailto:jhernand@redhat.com> > > <mailto:jhernand@redhat.com <mailto:jhernand@redhat.com>>> wrote: > > >>> > > >>> Hello, > > >>> > > >>> It has been recently requested that the API provides event > types: > > >>> > > >>> [RFE] Expose event types to API > > >>> https://bugzilla.redhat.com/1453170 > <https://bugzilla.redhat.com/1453170> > > <https://bugzilla.redhat.com/1453170 > <https://bugzilla.redhat.com/1453170>> > > >>> > > >>> Currently the API provides the event code and description, for > > example: > > >>> > > >>> <event href="/ovirt-engine/api/events/8021" id="8021"> > > >>> <code>19</code> > > >>> <description>Host myhost failed to recover.</description > > >>> ... > > >>> </event> > > >>> > > >>> There is no documentation of what is the meaning of codes, > > except the > > >>> source code of the engine itself. This forces some > applications > > to add > > >>> their own code to name mapping. For example, the 'ovirt' Ruby > > gem used > > >>> by older versions of ManageIQ to interact with oVirt contains > > the following: > > >>> > > >>> > > > https://github.com/ManageIQ/ovirt/blob/v0.17.0/lib/ovirt/ event.rb#L25-L485 > <https://github.com/ManageIQ/ovirt/blob/v0.17.0/lib/ovirt/ event.rb#L25-L485> > > > <https://github.com/ManageIQ/ovirt/blob/v0.17.0/lib/ovirt/ event.rb#L25-L485 > <https://github.com/ManageIQ/ovirt/blob/v0.17.0/lib/ovirt/ event.rb#L25-L485>> > > >>> > > >>> We could avoid this by adding to the API a new event > attribute that > > >>> indicates the type: > > >>> > > >>> <event href="/ovirt-engine/api/events/8021" id="8021"> > > >>> <code>19</code> > > >>> <type>host_recover_failure</type> > > >>> <description>Host myhost failed to recover.</description> > > >>> ... > > >>> </event> > > >>> > > >>> Ideally this should be defined as an enum, so that it will be > > >>> represented as an enum in the SDKs. Alternatively it could > just > > be an > > >>> string, and we could reuse the 'name' attribute: > > >>> > > >>> <event href="/ovirt-engine/api/events/8021" id="8021"> > > >>> <code>19</code> > > >>> <name>host_recover_failure</name> > > >>> <description>Host myhost failed to recover.</description> > > >>> ... > > >>> </event> > > >>> > > >>> However, the key point to making this useful would be to keep > > the types > > >>> (or names) backwards compatible, so that users of the API can > > rely on > > >>> their values and meanings. > > >>> > > >>> So this is my question to you: can we commit to keep the > names and > > >>> meanings of the backend event types backwards compatible? > > >> > > >> Do we even have to make it bw compatible? > > >> I guess it depends on the actual usage of those names… > > >> The ovirt ruby gem itself doesn’t do much with it > > > > > > We need to make keep it backwards compatible or else tell > users "don't > > > rely on these values, as they may change without notice". > > > > > > The 'ovirt' gem doesn't do anything special, it just creates > its own > > > code to name mapping. But the users of the 'ovirt' gem (the > ManageIQ > > > oVirt provider) do rely on the name. For example: > > > > > > > > > > https://github.com/ManageIQ/manageiq-providers-ovirt/blob/ master/app/models/manageiq/providers/redhat/infra_ manager/event_parser.rb#L80-L92 > <https://github.com/ManageIQ/manageiq-providers-ovirt/blob/ master/app/models/manageiq/providers/redhat/infra_ manager/event_parser.rb#L80-L92> > > > <https://github.com/ManageIQ/manageiq-providers-ovirt/blob/ master/app/models/manageiq/providers/redhat/infra_ manager/event_parser.rb#L80-L92 > <https://github.com/ManageIQ/manageiq-providers-ovirt/blob/ master/app/models/manageiq/providers/redhat/infra_ manager/event_parser.rb#L80-L92>> > > > > > > hmmm, while we are on topic, this pretty much looks like that manageiq > > does not only rely on the code but also on the actual value of it > since > > it is parsing it: > > > > # sample message: "Interface nic1 (VirtIO) was added to VM v5. (User: > > admin@internal-authz)" message.split(/\s/)[7][0...-1] > > > > Is this something we commit to maintain? Or should we commit to > maintain it? > > > > That is a good point, that isn't very future proof. We should also find > a way to make less fragile. Any suggestion? > > > The only doable thing which comes to my mind is something like this: > The msg is defined like this: > USER_ADD_VM_POOL_WITH_VMS_FAILED=Failed to create VM Pool ${VmPoolName} > (User: ${UserName}). > > e.g. the msg type and the variables. If we could expose in the api not > only the substituted msg but also the variable/value binding, we could > commit to keep the variable names backward compatible. > > So, something like: > > <event href="/ovirt-engine/api/events/8021" id="8021"> > <code>19</code> > <type>USER_ADD_VM_POOL_WITH_VMS_FAILED</type> > <description>the substituted msg.</description> > <parameters> > <parameter> > <key>VmPoolName</key> > <value>The Pool Name<value> > </parameter> > ... > </parameters> > </event> > > Not really rock solid since the variables would still be defined in the > AuditLogMessages.properties but still better and still easier to parse > on the client side. > That makes sense to me. I have opened the following bug to track that: [RFE] Add properties to events https://bugzilla.redhat.com/1456711 Note that for the particular case of the VM name, which is what ManageIQ is trying to do in that caode, the current best way is to use the <vm .../> link that is part of the event. I have opened the following ManageIQ issue to track it: Avoid parsing the descriptions of events https://github.com/ManageIQ/manageiq-providers-ovirt/issues/45 I think it was introduced while working on targeted refresh, but if we have an alternative then of course it would be better. > > > > > > > > > That means that if we ever change the meaning of a code the ManageIQ > > > provider, for example, will break. > > > > Right,then it indeed needs to stay stable. > > But how is maintaining the enum string different from the code? It is > > the same information, so if MIQ doesn't use the name directly then it > > doesn't really matter if it's a code or string. > > Perhaps deprecate the code and keep the name fixed? > > > > Thanks, > > michal > > > > > > > >>> > > >>> Regards, > > >>> Juan Hernandez > > >>> > > >>> > > >>> _______________________________________________ > > >>> Devel mailing list > > >>> Devel(a)ovirt.org <mailto:Devel@ovirt.org> > <mailto:Devel@ovirt.org <mailto:Devel@ovirt.org>> > > >>> http://lists.ovirt.org/mailman/listinfo/devel > <http://lists.ovirt.org/mailman/listinfo/devel> > > <http://lists.ovirt.org/mailman/listinfo/devel > <http://lists.ovirt.org/mailman/listinfo/devel>> > > > > > _______________________________________________ > > Devel mailing list > > Devel(a)ovirt.org <mailto:Devel@ovirt.org> > <mailto:Devel@ovirt.org <mailto:Devel@ovirt.org>> > > http://lists.ovirt.org/mailman/listinfo/devel > <http://lists.ovirt.org/mailman/listinfo/devel> > > <http://lists.ovirt.org/mailman/listinfo/devel > <http://lists.ovirt.org/mailman/listinfo/devel>> > > > > > > _______________________________________________ Devel mailing list Devel(a)ovirt.org http://lists.ovirt.org/mailman/listinfo/devel

2 1

[ OST Failure Report ] [ oVirt master ] [ 29/05/2017 ] [test_ovn_provider_keystone_token]
by Barak Korren 30 May '17

30 May '17

Test failed: [ test_ovn_provider_keystone_token ] Link to suspected patches: - https://gerrit.ovirt.org/#/c/76775/ Link to Job: - jenkins.ovirt.org/job/test-repo_ovirt_experimental_master/6899/ Link to all logs: - http://jenkins.ovirt.org/job/test-repo_ovirt_experimental_master/6899/artif… Error snippet from log: <error> Incorrect provider auth token value: PogQegXeOZwDg1dpKdrp-Wc2qJLt5QzSYMbeHZiPRU5wCCDOEaHRHXzvjY9FEso5Y2zdbzDIXHzxAT0XVpaKiQ -------------------- >> begin captured logging << -------------------- requests.packages.urllib3.connectionpool: INFO: * Starting new HTTPS connection (1): 192.168.201.3 py.warnings: WARNING: * Unverified HTTPS request is being made. Adding certificate verification is strongly advised. See: https://urllib3.readthedocs.org/en/latest/security.html requests.packages.urllib3.connectionpool: DEBUG: "POST /v2.0/tokens HTTP/1.1" 200 None --------------------- >> end captured logging << --------------------- </error> This had been failing constantly for 5 runs... -- Barak Korren RHV DevOps team , RHCE, RHCi Red Hat EMEA redhat.com | TRIED. TESTED. TRUSTED. | redhat.com/trusted

3 2