We've been fighting an extremely annoying issue where heavy Samba usage
causes a loss of network connectivity through an OSA card.  Sometimes
the problem will correct itself, like after 12 hours yesterday, and
sometimes we'll have to manually reload the qeth module.  Linux system
logs contain absolutely no errors.  The Samba logs report a connection
loss after the obligatory timeout, but nothing at the time of the
outage.  I'm now questioning whether we've got this device properly
defined.  Is 'portname' required for this to function correctly?  Where
are OSA errors logged to?  Some archived postings from 2002 mentioned
using addresses which start with denominations of 4 (i.e. 0, 4, 8), is
that still correct and a possible source of flaky behavior?


/etc/modules.conf
...
alias eth0 qeth
alias hsi0 qeth
...


/etc/chandev.conf
...
noauto;qeth0,0x0600,0x0601,0x0602
noauto;qeth1,0x0800,0x0801,0x0802


/var/log/boot.msg
...
<4>qeth: loading qeth S/390 OSA-Express driver ($Revision: 1.337.4.22
$/$Revision: 1.113.4.8 $/$Revision: 1.42.4.4 $:IPv6:VLAN)
<6> qeth: allocated 0 spare buffers
<4>qeth: Trying to use card with devnos 0x600/0x601/0x602
<4>qeth: Device 0x600/0x601/0x602 is an OSD Express card (level: 0551)
<4>with link type Gigabit Eth (portname:         )
<6> qeth: Broadcast packets will not be echoed back on eth0.
<4>qeth: Trying to use card with devnos 0x800/0x801/0x802
<4>qeth: Device 0x800/0x801/0x802 is a Guest LAN Hiper card (level:
V445)
<4>with link type magic
<4> qeth: IPv6 not supported on hsi0
...


Thanks.


Steve Kohrs

----------------------------------------------------------------------
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

Reply via email to