|
On Thu, Oct 28, 2010 at 01:52:15PM +0200, Hans Schillstrom wrote:
> Hello
>
> On Thursday 28 October 2010 13:35:12 Simon Horman wrote:
> > On Wed, Oct 27, 2010 at 10:58:31PM +0300, Julian Anastasov wrote:
> > >
> > > Hello,
> > >
> > > On Wed, 27 Oct 2010, Hans Schillstrom wrote:
> > >
> > > >>A New Spec of Type field:
> > > >>
> > > >>Bit 7 6 . . . 2 1 0
> > > >> +----------+--------------------------+-------------+-------+
> > > >> | Opt.Data | Spare | Packed IPv6 | IPv6 |
> > > >> +----------+--------------------------+-------------+-------+
> > > >
> > > >I can see a better usage of it in Option Type so Type will look like this
> > > > +-------------------------------------+-------------+-------+
> > > > | Spare | Packed IPv6 | IPv6 |
> > > > +-------------------------------------+-------------+-------+
> > > >
> > > >And "Option Type" in option field would look like this
> > > >
> > > >Bit 7 6 . . . 0 7 0
> > > > +----------+----------------------+---------------------------+
> > > > | Optional | Option type | Option length |
> > > > +----------+----------------------+---------------------------+
> >
> > As it stands a little more than 256 bytes may be needed for
> > pe_data (+ pe_name_length + pe_name). This could be resolved by
> > shortening the maximum pe_data length. Or perhaps we could use 16 bytes for
> > Option length, which should ensure its never too small.
> >
> > The 256 byte limit that I made for pe_data was arbitrarily chosen.
> >
>
> I have PE_NAME and PE_DATA ass different options
> so the limit is actually 255 bytes.
>
> #define IPVS_OPT_SEQ_DATA 1
> #define IPVS_OPT_PE_DATA 2
> #define IPVS_OPT_PE_NAME 3
>
> How ever they are not independant of each other.
> - PE_NAME never goes alone, only if there is PE_DATA.
> - In the receiving path, PE_NAME must have PE_DATA to be valid.
Ok, thanks, I haven't reviewed the patches that closely yet :-(
But still, 8 buts for option length does limit us to never ever
having an option longer than 255 bytes. But I guess we can
make sync protocol v3 if we need to cross that bridge.
> > > >We can have a better fine tuning of options in this way.
> > >
> > > Yes, that is exactly my idea. I more like the name
> > > "Parameter" instead of "Option", i.e. we have additional
> > > parameters that can be mandatory (usually) but also can be
> > > optional. For now I don't have idea for any optional
> > > parameters but allocating 1 bit for this does not look
> > > fatal.
> >
> > I'm not sure I understand the motivation for optional parameters.
> > I think its important to allow for backwards compatibility. But
> > I don't see that there will be multiple independent implementations
> > of the synchronisation daemon in the near future. So the use-case
> > isn't clear to me.
>
> Backward compatibility is "one way only"
> a new backup daemon can listen to an old one
> not the other way around.
Yes. I thought that was my point :-)
--
To unsubscribe from this list: send the line "unsubscribe lvs-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
|
