Hi Fujita-San > With the current path, still Dragonflow developers need to change > their code? If so, Muraoka-san, please fix the patch. If not, I'll > push this patch. >
For now, this patch does not support the number of non-zero. Therefore, to fix the patch. Also, add the following sentence to the document. If you set number to zone_src, Traceback occurs when you run the to_jsondict. Thanks, On 2016年07月27日 06:19, FUJITA Tomonori wrote: > On Tue, 26 Jul 2016 17:05:52 +0900 > IWAMOTO Toshihiro <iwam...@valinux.co.jp> wrote: > >> At Tue, 26 Jul 2016 16:25:15 +0900, >> Shinpei Muraoka wrote: >>> >>> Hi Iwamoto-San >>> >>> >>> On 2016年07月26日 13:55, IWAMOTO Toshihiro wrote: >>>> Thanks for cleaning up API incoherency. >>>> >>>> At Fri, 22 Jul 2016 13:59:49 +0900, >>>> Shinpei Muraoka wrote: >>>>> >>>>> Since zone_src in NXActionCT was not possible to specify string, >>>>> update zone_src of NXActionCT for the uniformity. >>>>> Therefore, you will be able to specify string of OXM/NXM fields for >>>>> zone_src. >>>> >>>> dragonflow is using number for zone_src now. >>>> Please consider allowing number for one release. >>>> >>>>> If you want to set the immediate value for zone, >>>>> zone_src set the None or empty character string. >>>> >>>> I don't think allowing an empty string here is a good idea. >>>> >>>> Using the predicate >>>> >>>> if zone_src is None: >>>> >>>> or >>>> >>>> if zone_src is None or zone_src == 0: >>>> >>>> (for backward compatibility) should be fine. >>>> >>> >>> I changed zone_src to string to keep to_jsondict method simple. >>> If you do not use jsondict, there is no problem by specifying a zero >>> to zone_src. >>> However, if you want to use parser, zone_src will be returned as empty >>> string. >>> dragonflow does not seem to use to_jsondict and parser. >>> For this reason, I think that there is no problem with the current >>> implementation. >> >> Ok. I see your point. >>> >>> The current zone_src can not allow a number other than zero. >>> How about correcting the following constant of dragonflow? >>> https://github.com/openstack/dragonflow/blob/master/dragonflow/controller/common/constants.py#L57-L59 >>> >>> $ git diff >>> diff --git a/dragonflow/controller/common/constants.py >>> b/dragonflow/controller/common/constants.py >>> index 106a1bf..796bf4b 100644 >>> --- a/dragonflow/controller/common/constants.py >>> +++ b/dragonflow/controller/common/constants.py >>> @@ -55,8 +55,8 @@ CT_STATE_TRK = 0x20 >>> CT_FLAG_COMMIT = 1 >>> >>> # Register match >>> -METADATA_REG = 0x80000408 >>> -CT_ZONE_REG = 0x1d402 >>> +METADATA_REG = 'metadata' >>> +CT_ZONE_REG = 'ct_zone' >>> >> >> This kind of change will be eventually needed, but if ryu suddenly >> stops to allow numeric zone_src, we'll need to coordinate with >> dragonflow people and schedule global-requirements change to avoid >> their bad experience. > > The whole point is that we can't ask others to change their code for > us randomly. So we have to preserve the previous API. > > With the current path, still Dragonflow developers need to change > their code? If so, Muraoka-san, please fix the patch. If not, I'll > push this patch. > ------------------------------------------------------------------------------ What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports.http://sdm.link/zohodev2dev _______________________________________________ Ryu-devel mailing list Ryu-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/ryu-devel
