Re: [Nbd] client-server via non TCP socket
- To: Paul Clements <paul.clements@...856...>
- Cc: firstname.lastname@example.org
- Subject: Re: [Nbd] client-server via non TCP socket
- From: Wouter Verhelst <w@...112...>
- Date: Sun, 6 May 2012 23:24:48 +0200
- Message-id: <20120506212448.GA10190@...3...>
- In-reply-to: <CAECXXi4VOObGXDwHaAj=GDXyPqe_cP86WYuNT+9wZbFrUZwDUg@...18...>
- References: <20120504141029.GA8845@...1102...> <87r4uy6dhp.fsf@...860...> <CAECXXi4VOObGXDwHaAj=GDXyPqe_cP86WYuNT+9wZbFrUZwDUg@...18...>
On Sun, May 06, 2012 at 01:17:43AM -0400, Paul Clements wrote:
> On Sat, May 5, 2012 at 8:04 AM, Goswin von Brederlow <goswin-v-b@...186...> wrote:
> > You will have to patch the nbd-client to initiate the ssh connection to
> > start the server instead of using socat or use some LD_PRELOAD lib that
> > changes socket() calls into opening a ssh connection. One problem there
> > might be that the kernel nbd module might expect a socket.
> Indeed, it does. It's fairly hardwired to having a socket for the connection.
> Can an stunnel or something like that not work?
It should; I've received many reports from people doing just that.
You shouldn't try to use it when swapping to NBD or similar (you'll
invite deadlocks that way), but for this use case it should not be any
less safe than 'regular' NBD.
The volume of a pizza of thickness a and radius z can be described by
the following formula:
pi zz a