On Thu, Apr 21, 2016 at 02:25:28PM -0400, Jan Harkes wrote:
> On Thu, Apr 21, 2016 at 02:16:02PM +0200, u-myfx_at_aetey.se wrote:
> > The only problem we hit there was not with clog but with memory alignment
> > in cfs (which we then fixed).
> 
> Looking forward to seeing patches related to that.

Attaching.

> > There is no http involved (it would be a large overhead without any reason).
> > 
> > This data does not have to be secured against MitM, the worst which can
> > happen is disruption of communication / DOS which MitM always can achieve.
> 
> Depends on what the configuration contains.

Correct. That's why we can not come to a conclusion without looking at
the actual code / specification. We can return to the subject when we
look at the corresponding patch later.

> > With ViceGetVolumeLocation() we make a similar mistake (!), we do
> > not allow for the returned string to be invalidated "properly". There
> > should have been some kind of a validity promise (say a TTL) included.
> > A possibility to change host names in a realm setup without confusing
> > the clients is a Good Thing (TM), why forbid this by design?
> 
> GetVolumeLocation returns a DNS hostname, DNS records have a TTL. There
> is nothing here forbidden by design, we just don't duplicate something
> that is already there.

In your scheme there are several mappings to do:
 volume/server -[1]-> dnsname -[2]-> ip

It was not about [2] but about [1] which "risks to fail to become invalidated"
when it is stale.

I understand now that you implied to reresolve volume/server->dnsname
when the client loses contact with a server.

We in our code trigger reresolution when a server becomes unreachable
(for "too long").

Then indeed no special expiration time is needed.

> It just seems like we're having the same discussions over and over

You are right, we have already presented the arguments a couple
of years ago and also commented on them, no need to duplicate.

> again, and yet it always results in to us disagreeing and you just

No, not _that_ bad. We actually agree on most things, we just put the
discussion effort into the ones where we do not :) the other ones go
unnoticed.

Think also, if I wouldn't value your comments, I wouldn't ask for them.

> doing your thing anyway,

That's how the things get done. One implements what one feels is the
most suitable approach.

> Jan

Best regards,
Rune