[nanomsg] Re: NN_SNDBUF on inproc
- From: Martin Sustrik <sustrik@xxxxxxxxxx>
- To: nanomsg@xxxxxxxxxxxxx
- Date: Wed, 08 Oct 2014 22:46:54 +0200
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 08/10/14 21:27, Steve Vinoski wrote:
> Hi Martin, you're right that RCVBUF is in control in this context.
> The problem with that, though, is that it prevents setting the
> buffer size if you try to set it on socket that doesn't support
> RCVBUF, such as a push socket for example.
I am not 100% sure how that works on the kernel level (any insights,
Garrett?) but my understanding is that owner of the IPC socket can
only specify the size of its rx buffer and has no way to affect the
size of its tx buffer. Tx buffer size is specified by the owner of the
peer socket -- from whose point of view it is an rx buffer.
So, if you have an unidirectional send-only socket, such as PUSH,
there's nothing you can do about the buffer sizes: rx buffer is unused
anyway and thus setting its size is irrelevant and tx buffer is set by
the peer.
Am I missing something here?
Martin
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
iQEcBAEBAgAGBQJUNaK2AAoJENTpVjxCNN9YGlcH/iDrk8uFTmd4ezWloX/Sgk2k
s5mKbohQvdcpgZSYFRBOYvBEZaFnDmKNxUegGWJKXdsJ+tiyV+sXoaeP9W6RHbU6
NCZRqqwbCcb6ivFRCrXBjfCw1suqNhoYByb9y2XJuGPZCRAvyTtdn3iqyG1kssLZ
tazsBoqS9uve75aS4A3C+a1nJ7X9wWIZ8GatOAjlKjbDvClA4t0yDuagTeqZlUlY
0LNuzQqXbP0YViR8e4NJDByEMvPf1SlIZGpLYkuxhFfG4NnhQ6AB+w5EAgWnlGY5
7XAjg1mTc874sHtJsrmyQsg5VLaeYQdAMtwNvgA9wH+2uELjoCiPtalOaROYCvE=
=oSmr
-----END PGP SIGNATURE-----
Other related posts:
