How do you set the MTU so high? I'm trying to play with super jumbo frames on a SLES10 guest under Z/VM and it doesnt appear to allow an MTU value larger than 12256:
db1:~ # ifconfig hsi0 mtu 12257 SIOCSIFMTU: Invalid argument db1:~ # ifconfig hsi0 mtu 12256 db1:~ # ifconfig hsi0 mtu 16384 SIOCSIFMTU: Invalid argument What did you have to do to get yours to accept 16384? -----Original Message----- From: Martin, Terry R. (CMS/CTR) (CTR) <[EMAIL PROTECTED]> To: LINUX-390@VM.MARIST.EDU Sent: Tue, 19 Aug 2008 12:01 pm Subject: Re: Help with network settings on z/Linux Hi Alan, I did some research based on your response I this is what I found: In my case the E3 (HiperSockets CHPID) with a CHPARM=40k is being used in the environment that is working. The E2 (HiperSockets CHPID) with a CHPARM=64k is being used in the environment that is NOT working. Can the difference in these sizes alone cause me the issue I am having? The MTU size for both HiperSockets is set to 16,348 across the board. Thank You, Terry Martin Lockheed Martin - Information Technology z/OS & z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] -----Original Message----- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Alan Altmark Sent: Tuesday, August 19, 2008 11:10 AM To: LINUX-390@VM.MARIST.EDU Subject: Re: Help with network settings on z/Linux On Tuesday, 08/19/2008 at 10:41 EDT, "Martin, Terry R. (CMS/CTR) (CTR)" <[EMAIL PROTECTED]> wrote:
We have two environments that we are building as part of our POC. We ship data via DB2 Connect from the mainframe DB2 to a Linux guest via
a
HiperSockets Network. In our DEV environment all of our tables process without issue. In our VAL environment there are six tables that the
SQL
fails for. We have been engaged with IBM via a PMR on this and have
been
going back and forth. At this point it appears that the issue is with the number of rows. In the SQL when they do an SELECT * and do not specify the rows it works. When they specify the tables in the Select
it
apparently increases the size of the SQL and it never gets to the mainframe. It looks like it is a 2k buffer limitation somewhere. Is there anything in the TCP/IP stack in z/Linux or any network setting
in
z/Linux that would enforce this limitation?
Verify that the MFS (CHPARM= or OS= in IOCP) and MTU for the HiperSockets connection are the same for your DEV and VAL environments. Alan Altmark z/VM Development IBM Endicott ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390