Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-06-09 Thread Kim Kimball




Jeffrey Altman wrote:

  Kim Kimball wrote:
  
  
Douglas E. Engert wrote:


  
Kim Kimball wrote:
  
  
Have you tried setting MTU on the fileserver command line instead?

  
  No, and I don't want to. Our problem is with just a few clients using
VPNs
I don't want to slow everyone else down.

  

Didn't for us, couldn't be measured to any significance.

YMMV

  
  
Kim:

Your issue was slightly different in that in your case you had
old Transarc clients that could not be modified in any way.
A server only solution was the only thing that could work.

Jeffrey Altman
  


Hi Jeffrey,

Thanks for the reminder.  I'd forgotten.  And it's still the case that
we're running antique clients, probably forever or until ...

Kim





___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-29 Thread Jeffrey Altman
Kim Kimball wrote:
> Douglas E. Engert wrote:
>>
>>
>> Kim Kimball wrote:
>>> Have you tried setting MTU on the fileserver command line instead?
>>
>> No, and I don't want to. Our problem is with just a few clients using
>> VPNs
>> I don't want to slow everyone else down.
>>
> Didn't for us, couldn't be measured to any significance.
> 
> YMMV

Kim:

Your issue was slightly different in that in your case you had
old Transarc clients that could not be modified in any way.
A server only solution was the only thing that could work.

Jeffrey Altman


smime.p7s
Description: S/MIME Cryptographic Signature


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-29 Thread Kim Kimball

Douglas E. Engert wrote:



Kim Kimball wrote:

Have you tried setting MTU on the fileserver command line instead?


No, and I don't want to. Our problem is with just a few clients using 
VPNs

I don't want to slow everyone else down.


Didn't for us, couldn't be measured to any significance.

YMMV



That's what we've done, for the same reason.  It has worked, and we 
don't have to fiddle with each client.  The negotiation between AFS 
client and fileserver for MTU size is "fileserver wins."



I got a mod working over the weekend, on Ubuntu (openafs 1.4.7) and
MacOS 10.4 with OpenAFS 1.4.10, that adds a -rxmaxmtu option to afsd.
I am putting final touches on it today.


And we'll use it to good advantage, as it is of course a better solution 
for non VPN users, as you note.





Kim Kimball

Douglas E. Engert wrote:

We are having problems with Mac OS 10.4 and 10.5  using Cisco VPN
AFS can become unusable.  Mac 10.4 is running OpenAFS 1.4.8 for sure.
I think the Mac 10.5 is running OpenAFS 1.5.59.

Using rxdebug and looking at the natMTU parameter, on most Unix systems
this is 1444,(1500 - 56) as expected. On Windows systems this is 
usually

1260.  And on MAC it is 1444.

Even if I set the interface mtu 1244, and reboot the MAC, rxdebug shows
the interface is using 1244 but rxdebug continues to show a natMTU = 
1444.

as though it still assumed the mtu was 1500.

So it looks like the MAC client is not getting the existingMTU
from the OS in util/netutils.c

The AFS client on Windows has the RxMaxMTU (1244 appears to be the best
setting). Is there any equivelent option for the MAC?

Any thoughts on this?










___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-28 Thread Jeff Blaine

Jeff Blaine wrote:

FWIW, this appears to be the same problem I reported in
April, but for Windows.

https://lists.openafs.org/pipermail/openafs-info/2009-April/031127.html

We are still working with our networking+VPN folks to
try to determine if it's the same thing or not, as well
as how to fix it.



Have you tried the Windows registry setting rxMaxMtu?

It needs to be 56 bytes less then the actually MTU.
Cisco VPN appears to force a MTU of 1300 on the interface,
so rxMaxMtu should be 1244.


Setting it to 1244 or lower did in fact cause it to start
working.

Well, it works as long as CheckPoint Integrity Flex is
completely off.  With it on, even with all functionality
turned off (Firewall, Program Control, etc), AFS fails.
Figure that one out!  :|


If you are having a fragmentation problem, ping can help, by setting
the don't fragment bit and varying the size can help see what the
limit is.

On Windows:   ping -f -l  ...
On Linux: ping -M do -s  


On the windows client that is failing, try:
rxdebug 127.0.0.1 -port 7001 -peer -long
ipconfig /all

Look for the ifMTU  natMTU  maxMTU 
lines for each afs server. Also compare the IP numbers
with the ip number in the ipconfig.


Very helpful debugging notes.  Thanks for sharing!


As best as I can tell, The code in rx_kcommon.c tries to
determine the ifMTU based on the addresses of the client and server.
If they appear to be on the same class a, b, or c network, it
may work by using the MTU from the matching interface. But none
if the interfaces match at all, it defaults to using 1500, which can case
fragmentation. Even in the Windows case I think this might happen.

To test this I need to get off site. Will try tonight from home.

___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-28 Thread Jeff Blaine
"In order to enable OpenAFS to operate across releases of the Cisco 
IPSec VPN client prior than 5.0"


We're not running sub 5.0 Cisco VPN software, but I'll give it
a shot.

Jeffrey Altman wrote:

Jeff Blaine wrote:

FWIW, this appears to be the same problem I reported in
April, but for Windows.

https://lists.openafs.org/pipermail/openafs-info/2009-April/031127.html

We are still working with our networking+VPN folks to
try to determine if it's the same thing or not, as well
as how to fix it.


Read the Release Notes

http://docs.openafs.org/ReleaseNotesWindows/ch08.html#id4834377

Search for RxMaxMTU or Cisco.


___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-28 Thread Jeffrey Altman
Jeff Blaine wrote:
> FWIW, this appears to be the same problem I reported in
> April, but for Windows.
> 
> https://lists.openafs.org/pipermail/openafs-info/2009-April/031127.html
> 
> We are still working with our networking+VPN folks to
> try to determine if it's the same thing or not, as well
> as how to fix it.

Read the Release Notes

http://docs.openafs.org/ReleaseNotesWindows/ch08.html#id4834377

Search for RxMaxMTU or Cisco.



smime.p7s
Description: S/MIME Cryptographic Signature


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-28 Thread Douglas E. Engert



Jeff Blaine wrote:

FWIW, this appears to be the same problem I reported in
April, but for Windows.

https://lists.openafs.org/pipermail/openafs-info/2009-April/031127.html

We are still working with our networking+VPN folks to
try to determine if it's the same thing or not, as well
as how to fix it.



Have you tried the Windows registry setting rxMaxMtu?

It needs to be 56 bytes less then the actually MTU.
Cisco VPN appears to force a MTU of 1300 on the interface,
so rxMaxMtu should be 1244.

If you are having a fragmentation problem, ping can help, by setting
the don't fragment bit and varying the size can help see what the
limit is.

On Windows:   ping -f -l  ...
On Linux: ping -M do -s  


On the windows client that is failing, try:
rxdebug 127.0.0.1 -port 7001 -peer -long
ipconfig /all

Look for the ifMTU  natMTU  maxMTU 
lines for each afs server. Also compare the IP numbers
with the ip number in the ipconfig.

As best as I can tell, The code in rx_kcommon.c tries to
determine the ifMTU based on the addresses of the client and server.
If they appear to be on the same class a, b, or c network, it
may work by using the MTU from the matching interface. But none
if the interfaces match at all, it defaults to using 1500, which can case
fragmentation. Even in the Windows case I think this might happen.

To test this I need to get off site. Will try tonight from home.





___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info




--

 Douglas E. Engert  
 Argonne National Laboratory
 9700 South Cass Avenue
 Argonne, Illinois  60439
 (630) 252-5444
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-28 Thread Jeff Blaine

FWIW, this appears to be the same problem I reported in
April, but for Windows.

https://lists.openafs.org/pipermail/openafs-info/2009-April/031127.html

We are still working with our networking+VPN folks to
try to determine if it's the same thing or not, as well
as how to fix it.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-27 Thread Douglas E. Engert


For anyone interested, I sent in a patch today for this
[grand.central.org #124880] AutoReply: Add -rxmaxmtu parameter to afsd to help 
avoid VPN fragmentation of UDP packets

Kim Kimball wrote:

Have you tried setting MTU on the fileserver command line instead?

That's what we've done, for the same reason.  It has worked, and we 
don't have to fiddle with each client.  The negotiation between AFS 
client and fileserver for MTU size is "fileserver wins."


Kim Kimball

Douglas E. Engert wrote:

We are having problems with Mac OS 10.4 and 10.5  using Cisco VPN
AFS can become unusable.  Mac 10.4 is running OpenAFS 1.4.8 for sure.
I think the Mac 10.5 is running OpenAFS 1.5.59.

Using rxdebug and looking at the natMTU parameter, on most Unix systems
this is 1444,(1500 - 56) as expected. On Windows systems this is usually
1260.  And on MAC it is 1444.

Even if I set the interface mtu 1244, and reboot the MAC, rxdebug shows
the interface is using 1244 but rxdebug continues to show a natMTU = 
1444.

as though it still assumed the mtu was 1500.

So it looks like the MAC client is not getting the existingMTU
from the OS in util/netutils.c

The AFS client on Windows has the RxMaxMTU (1244 appears to be the best
setting). Is there any equivelent option for the MAC?

Any thoughts on this?







--

 Douglas E. Engert  
 Argonne National Laboratory
 9700 South Cass Avenue
 Argonne, Illinois  60439
 (630) 252-5444
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-27 Thread Douglas E. Engert



Kim Kimball wrote:

Have you tried setting MTU on the fileserver command line instead?


No, and I don't want to. Our problem is with just a few clients using VPNs
I don't want to slow everyone else down.



That's what we've done, for the same reason.  It has worked, and we 
don't have to fiddle with each client.  The negotiation between AFS 
client and fileserver for MTU size is "fileserver wins."



I got a mod working over the weekend, on Ubuntu (openafs 1.4.7) and
MacOS 10.4 with OpenAFS 1.4.10, that adds a -rxmaxmtu option to afsd.
I am putting final touches on it today.




Kim Kimball

Douglas E. Engert wrote:

We are having problems with Mac OS 10.4 and 10.5  using Cisco VPN
AFS can become unusable.  Mac 10.4 is running OpenAFS 1.4.8 for sure.
I think the Mac 10.5 is running OpenAFS 1.5.59.

Using rxdebug and looking at the natMTU parameter, on most Unix systems
this is 1444,(1500 - 56) as expected. On Windows systems this is usually
1260.  And on MAC it is 1444.

Even if I set the interface mtu 1244, and reboot the MAC, rxdebug shows
the interface is using 1244 but rxdebug continues to show a natMTU = 
1444.

as though it still assumed the mtu was 1500.

So it looks like the MAC client is not getting the existingMTU
from the OS in util/netutils.c

The AFS client on Windows has the RxMaxMTU (1244 appears to be the best
setting). Is there any equivelent option for the MAC?

Any thoughts on this?







--

 Douglas E. Engert  
 Argonne National Laboratory
 9700 South Cass Avenue
 Argonne, Illinois  60439
 (630) 252-5444
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


Re: [OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-26 Thread Kim Kimball

Have you tried setting MTU on the fileserver command line instead?

That's what we've done, for the same reason.  It has worked, and we 
don't have to fiddle with each client.  The negotiation between AFS 
client and fileserver for MTU size is "fileserver wins."


Kim Kimball

Douglas E. Engert wrote:

We are having problems with Mac OS 10.4 and 10.5  using Cisco VPN
AFS can become unusable.  Mac 10.4 is running OpenAFS 1.4.8 for sure.
I think the Mac 10.5 is running OpenAFS 1.5.59.

Using rxdebug and looking at the natMTU parameter, on most Unix systems
this is 1444,(1500 - 56) as expected. On Windows systems this is usually
1260.  And on MAC it is 1444.

Even if I set the interface mtu 1244, and reboot the MAC, rxdebug shows
the interface is using 1244 but rxdebug continues to show a natMTU = 
1444.

as though it still assumed the mtu was 1500.

So it looks like the MAC client is not getting the existingMTU
from the OS in util/netutils.c

The AFS client on Windows has the RxMaxMTU (1244 appears to be the best
setting). Is there any equivelent option for the MAC?

Any thoughts on this?




___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info


[OpenAFS] OpenAFS, Cisco VPN and MAC OS and mtu

2009-05-15 Thread Douglas E. Engert

We are having problems with Mac OS 10.4 and 10.5  using Cisco VPN
AFS can become unusable.  Mac 10.4 is running OpenAFS 1.4.8 for sure.
I think the Mac 10.5 is running OpenAFS 1.5.59.

Using rxdebug and looking at the natMTU parameter, on most Unix systems
this is 1444,(1500 - 56) as expected. On Windows systems this is usually
1260.  And on MAC it is 1444.

Even if I set the interface mtu 1244, and reboot the MAC, rxdebug shows
the interface is using 1244 but rxdebug continues to show a natMTU = 1444.
as though it still assumed the mtu was 1500.

So it looks like the MAC client is not getting the existingMTU
from the OS in util/netutils.c

The AFS client on Windows has the RxMaxMTU (1244 appears to be the best
setting). Is there any equivelent option for the MAC?

Any thoughts on this?

--

 Douglas E. Engert  
 Argonne National Laboratory
 9700 South Cass Avenue
 Argonne, Illinois  60439
 (630) 252-5444
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info