[ovirt-devel] Making event types backwards compatible?
Moti Asayag
masayag at redhat.com
Wed Jun 7 07:40:08 UTC 2017
On Tue, May 30, 2017 at 11:28 AM, Juan Hernández <jhernand at 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 at redhat.com
> > <mailto:jhernand at 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 at redhat.com <mailto:mskrivan at redhat.com>
> > > <mailto:mskrivan at redhat.com <mailto:mskrivan at redhat.com>>> wrote:
> > >
> > > > On 29 May 2017, at 11:44, Juan Hernández <
> jhernand at redhat.com <mailto:jhernand at redhat.com>
> > <mailto:jhernand at redhat.com <mailto:jhernand at redhat.com>>> wrote:
> > > >
> > > >> On 05/29/2017 11:27 AM, Michal Skrivanek wrote:
> > > >>
> > > >>> On 29 May 2017, at 10:39, Juan Hernández <
> jhernand at redhat.com <mailto:jhernand at redhat.com>
> > > <mailto:jhernand at redhat.com <mailto:jhernand at 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?
>
we can maintain the mapping layer between backend event types to restapi
event types
on the restapi side. This is currently done on the ovirt-gem, and once
ovirt-gem is depracated
we could rely the manageiq-providers-ovirt code on the api/sdk which will
expose the event name.
Having a single api maintainer will ease the maintenance of it and will
allow a better control of
changes to that part.
> > > >>
> > > >> 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 at 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
>
>
I don't think we should expose the internal implementation of the
audit-logging mechanism via the api.
This is unreliable and not the recommended way IMO to obtain data related
to the event.
If the concern is for the entities names, we can extend the returned event
element in the response
to include also the name:
<event>
<vm id="..." href="...">
<name>abc</name>
</vm>
This can be applied to all of the main entities without penalty in data
retrieval from the database, since
the names of the entities should also be available on the audit-log entity.
However, for all of the other placesholders - this is a coupling of the
restapi event structure to the
implementation of the auditlogging:
With the current code we don't store as part of the event data all of the
raw material of the event,
specifically any variable being set via the AuditLogable.#customValues().
Therefore some of the data won't be available for the returned event.
> Note that for the particular case of the VM name, which is what ManageIQ
> is trying to do in that code, 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
>
> >
> > >
> > > >
> > > > 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 at ovirt.org <mailto:Devel at ovirt.org>
> > <mailto:Devel at ovirt.org <mailto:Devel at 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 at ovirt.org <mailto:Devel at ovirt.org>
> > <mailto:Devel at ovirt.org <mailto:Devel at 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 at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/devel
--
Regards,
Moti
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/devel/attachments/20170607/458f2303/attachment-0001.html>
More information about the Devel
mailing list