Hi Nick,
I have updated the patch as you suggested. I have kept
InitializeMTUPropertyValue() method because global MTU value can be
initialized only once.
Venkat
Nick Wagner <ne...@wingedbeast.org>
Sent by: nwagner...@gmail.com
23/08/2010 20:26
To
Venkateswara R Puvvada/India/i...@ibmin
cc
Madhuri D Madhavanpillai/India/i...@ibmin,
openslp-devel@lists.sourceforge.net, "Morrell Richard (external)"
<richard.morr...@uk.thalesgroup.com>, Varun Chandramohan
<var...@linux.vnet.ibm.com>
Subject
Re: [Openslp-devel] Regd: Openslp MTU issue
Perhaps, instead of even having InitializeMTUPropertyValue(), change
SLPNetworkSetSendRecvBuf to getsockopt, see if it's greater than
global.mtu, and if not setsockopt to global.mtu. That would provide the
minimum change (and therefore minimum new side effects) to how the sockets
work.
If we must keep InitializeMTUPropertyValue, it should do a getsockopt and
compare before setting the MTU. If the socket option is larger than
global.MTU, I would prefer that we don't actually change the socket
options. The existing code doesn't touch the socket buffer sizes, and it
is unknown what side effects will happen if we start limiting them. And
the default receive buffer size is often larger than the default send
buffer, so limiting them both to the same value could also have side
effects.
Thanks for looking at this!
--Nick
On Mon, Aug 23, 2010 at 6:36 AM, Venkateswara R Puvvada <
vpuvv...@in.ibm.com> wrote:
Hi Nick,
I have incorporated your review comments and uploaded the patch to the
bugzilla. Please let me know if you have any further comments.
Venkat
Nick Wagner <ne...@wingedbeast.org>
Sent by: nwagner...@gmail.com
21/08/2010 02:47
To
Varun Chandramohan <var...@linux.vnet.ibm.com>
cc
openslp-devel@lists.sourceforge.net, "Morrell Richard (external)" <
richard.morr...@uk.thalesgroup.com>, Venkateswara R
Puvvada/India/i...@ibmin, Madhuri D Madhavanpillai/India/i...@ibmin
Subject
Re: [Openslp-devel] Regd: Openslp MTU issue
I think a similar change could be made to the slpd_socket.c, and I would
check the other areas that call socket(). Instead of just setting the
options, I would also do a get and compare with mtu. Some OSs have quite
large values for those options already (sometimes greater than 65535), no
need to shrink them down.
--Nick
2010/8/20 Varun Chandramohan <var...@linux.vnet.ibm.com>
Hi Nick,
The patch is updated in the tracker, can you have a look at
this? Let us know if its ok?
Regards,
Varun
On Friday, August 13, 2010 09:54:48 am Varun Chandramohan wrote:
> Hi Nick,
>
> I think you are right. May be we should have a check for
that. Iam adding the author of the patch to this mail. He is not
subscribed to this ML, so please do reply all
> so that he is kept in loop. In our labs we have had major issues to get
this working. If you have suggestion do let us know so that we can address
this at the earliest.
>
> Venkat,
> Can you look into this issue?
>
> Regards,
> Varun
> On Friday, August 13, 2010 12:32:46 am Nick Wagner wrote:
> > Won't we need to make sure the socket options SO_SNDBUF and SO_RCVBUF
fit
> > MTU? Doing a UDP sendto, for example can cause an EMSGSIZE if the
message
> > is longer than SO_SNDBUF.
> >
> > --Nick
> >
> > On Wed, Aug 4, 2010 at 10:07 PM, Varun Chandramohan <
> > var...@linux.vnet.ibm.com> wrote:
> >
> > > Hi Richard,
> > >
> > > As per the below discussion Venkat had updated the patch. Can you
please
> > > take a look and commit these changes?
> > >
> > >
> > >
http://sourceforge.net/tracker/?func=detail&aid=3016317&group_id=1730&atid=101730
> > >
> > > Regards,
> > >
> > > Varun
> > >
> > > On Thursday, July 08, 2010 02:11:49 pm Morrell Richard (external)
wrote:
> > >
> > > > In the daemon code, I would suggest we store the MTU in the
> > > G_SlpdProperty
> > >
> > > > structure when the config file is read or re-read (as in Venkat
Puvvada's
> > >
> > > > patch), and replace all references to SLPD_MAX_DATAGRAM_SIZE by
> > >
> > > > G_SlpdProperty.MTU throughout the daemon source.
> > >
> > > >
> > >
> > > > In the library code, there is (currently) no global property
structure.
> > > We
> > >
> > > > could introduce one, expanding libslp_property.c to initialise it,
or
> > > just
> > >
> > > > initialise a local variable when needed by calling
> > >
> > > > SLPPropertyAsInteger("net.slp.MTU"), and replace the references to
> > >
> > > > SLPD_MAX_DATAGRAM_SIZE by a reference to the local variable (as is
> > > partially
> > >
> > > > done in libslp_network.c).
> > >
> > > >
> > >
> > > > The SLPD_MAX_DATAGRAM_SIZE macro should also be removed.
> > >
> > > >
> > >
> > > > -----Original Message-----
> > >
> > > > From: Varun Chandramohan [mailto:var...@linux.vnet.ibm.com]
> > >
> > > > Sent: 08 July 2010 09:17
> > >
> > > > To: Morrell Richard (external)
> > >
> > > > Cc: 'openslp-devel@lists.sourceforge.net'; liwei...@cn.ibm.com;
> > >
> > > > viksa...@in.ibm.com; vpuvv...@in.ibm.com
> > >
> > > > Subject: Re: Regd: Openslp MTU issue
> > >
> > > >
> > >
> > > >
> > >
> > > > I agree, we need a better solution to the problem,i think
Venkatesh got
> > > it
> > >
> > > > right, but its not done fully. May be if you can point out the
possible
> > >
> > > > solution you are looking for, he might be able to redo
> > >
> > > > the patch and submit.
> > >
> > > >
> > >
> > > > Wei Wei, Venkateswara, and Vikram can you guys discuss here and
take it
> > >
> > > > forward?
> > >
> > > >
> > >
> > > > Regards,
> > >
> > > > Varun
> > >
> > > > On Thursday, July 08, 2010 01:19:57 pm Morrell Richard (external)
wrote:
> > >
> > > > > Yes, we had a project that was registering CORBA IORs that ran
to over
> > >
> > > > 3000
> > >
> > > > > bytes, and had a similar problem.
> > >
> > > > >
> > >
> > > > > Someone suggested, as a "quick fix" that we change the
> > >
> > > > SLP_MAX_DATAGRAM_SIZE
> > >
> > > > > macro as follows:
> > >
> > > > > #define SLP_MAX_DATAGRAM_SIZE
SLPPropertyAsInteger("net.slp.MTU")
> > >
> > > > > so we could control the MTU via the config file.
> > >
> > > > >
> > >
> > > > > As the macro is never used in a compile-time only context (eg.
as the
> > > size
> > >
> > > > > of an array), and the SLPPropertyAsInteger call works in both
the
> > > client
> > >
> > > > > library and the daemon code, this worked for that situation.
However, I
> > >
> > > > > wasn't happy about it in a general sense, so I didn't make the
change
> > > in
> > >
> > > > the
> > >
> > > > > OpenSLP code (we apply it as a patch when building). I haven't
had the
> > >
> > > > time
> > >
> > > > > to do a proper job along the lines of that proposed by Venkat
Puvvada
> > > (but
> > >
> > > > > extended to cover all the references to SLP_MAX_DATAGRAM_SIZE in
both
> > >
> > > > daemon
> > >
> > > > > and library code).
> > >
> > > > >
> > >
> > > > > We would certainly support such a fix.
> > >
> > > > >
> > >
> > > > > Regards,
> > >
> > > > >
> > >
> > > > > Richard Morrell
> > >
> > > > >
> > >
> > > > > -----Original Message-----
> > >
> > > > > From: Varun Chandramohan [mailto:var...@linux.vnet.ibm.com]
> > >
> > > > > Sent: 07 July 2010 20:59
> > >
> > > > > To: Morrell Richard (external)
> > >
> > > > > Cc: 'openslp-devel@lists.sourceforge.net'; John Calcote
> > >
> > > > > Subject: Fwd: Re: Regd: Openslp MTU issue
> > >
> > > > >
> > >
> > > > >
> > >
> > > > > To take is discussion forward. This sounds like a important
issue to me
> > >
> > > > and
> > >
> > > > > we can work on a fix if the fix currently provided is right
approach.
> > >
> > > > >
> > >
> > > > > Richard, any thoughts?
> > >
> > > > >
> > >
> > > > > This email, including any attachment, is a confidential
communication
> > >
> > > > > intended solely for the use of the individual or entity to whom
it is
> > >
> > > > > addressed. It contains information which is private and may be
> > > proprietary
> > >
> > > > > or covered by legal professional privilege. If you have received
this
> > >
> > > > email
> > >
> > > > > in error, please notify the sender upon receipt, and immediately
delete
> > > it
> > >
> > > > > from your system.
> > >
> > > > >
> > >
> > > > > Anything contained in this email that is not connected with the
> > > businesses
> > >
> > > > > of this company is neither endorsed by nor is the liability of
this
> > >
> > > > company.
> > >
> > > > >
> > >
> > > > > Whilst we have taken reasonable precautions to ensure that any
> > > attachment
> > >
> > > > to
> > >
> > > > > this email has been swept for viruses, we cannot accept
liability for
> > > any
> > >
> > > > > damage sustained as a result of software viruses, and would
advise that
> > >
> > > > you
> > >
> > > > > carry out your own virus checks before opening any attachment.
> > >
> > > > >
> > >
> > > > >
> > >
> > > >
> > >
> > > > This email, including any attachment, is a confidential
communication
> > >
> > > > intended solely for the use of the individual or entity to whom it
is
> > >
> > > > addressed. It contains information which is private and may be
> > > proprietary
> > >
> > > > or covered by legal professional privilege. If you have received
this
> > > email
> > >
> > > > in error, please notify the sender upon receipt, and immediately
delete
> > > it
> > >
> > > > from your system.
> > >
> > > >
> > >
> > > > Anything contained in this email that is not connected with the
> > > businesses
> > >
> > > > of this company is neither endorsed by nor is the liability of
this
> > > company.
> > >
> > > >
> > >
> > > > Whilst we have taken reasonable precautions to ensure that any
attachment
> > > to
> > >
> > > > this email has been swept for viruses, we cannot accept liability
for any
> > >
> > > > damage sustained as a result of software viruses, and would advise
that
> > > you
> > >
> > > > carry out your own virus checks before opening any attachment.
> > >
> > > >
> > >
> > > >
> > >
> > > >
> > >
------------------------------------------------------------------------------
> > >
> > > > This SF.net email is sponsored by Sprint
> > >
> > > > What will you do first with EVO, the first 4G phone?
> > >
> > > > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> > >
> > > > _______________________________________________
> > >
> > > > Openslp-devel mailing list
> > >
> > > > Openslp-devel@lists.sourceforge.net
> > >
> > > > https://lists.sourceforge.net/lists/listinfo/openslp-devel
> > >
> > > >
> > >
> > >
> > >
> > >
------------------------------------------------------------------------------
> > > The Palm PDK Hot Apps Program offers developers who use the
> > > Plug-In Development Kit to bring their C/C++ apps to Palm for a
share
> > > of $1 Million in cash or HP Products. Visit us here for more
details:
> > > http://p.sf.net/sfu/dev2dev-palm
> > > _______________________________________________
> > > Openslp-devel mailing list
> > > Openslp-devel@lists.sourceforge.net
> > > https://lists.sourceforge.net/lists/listinfo/openslp-devel
> > >
> > >
> >
>
------------------------------------------------------------------------------
Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
Be part of this innovative community and reach millions of netbook users
worldwide. Take advantage of special opportunities to increase revenue and
speed time-to-market. Join now, and jumpstart your future.
http://p.sf.net/sfu/intel-atom-d2d
_______________________________________________
Openslp-devel mailing list
Openslp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openslp-devel
