[openfabrics-ewg] IPoIB HA not working on RHEL4 U3, please fix for OFED 1.1 rc7

[Scott Weitzenkamp (sweitzen)]

> -----Original Message-----
> From: bugzilla-daemon at openib.org [mailto:bugzilla-daemon at openib.org] 
> Sent: Monday, September 25, 2006 12:47 AM
> To: Scott Weitzenkamp (sweitzen)
> Subject: [Bug 247] New: OFED IPoIB HA not working on RHEL4 U3
> 
> http://openib.org/bugzilla/show_bug.cgi?id=247
> 
>            Summary: OFED IPoIB HA not working on RHEL4 U3
>            Product: OpenFabrics Linux
>            Version: 1.1rc5
>           Platform: All
>         OS/Version: RHEL 4
>             Status: NEW
>           Severity: critical
>           Priority: P2
>          Component: IPoIB
>         AssignedTo: vlad at mellanox.co.il
>         ReportedBy: sweitzen at cisco.com
> 
> 
> OFED 1.1 rc5 on RHEL4 U3 x86_64.
> 
> I cannot get IPoIB HA working.  I see ipoib_ha.pl uses 
> "/sbin/ip monitor link
> all" and looks for "NO-CARRIER".  I don't seem to be getting 
> this output on
> RHEL4 U3.   What operating system(s) has Mellanox tested IPoIB HA on?
> 
> I ran "/sbin/ip monitor link all" manually, and when I shut 
> down the ib0 IB
> port (from the switch), I see this:
> 
> 4: ib0: <BROADCAST,MULTICAST,UP> mtu 2044 qdisc pfifo_fast
>     link/[32] 
> 00:00:04:04:fe:80:00:00:00:00:00:00:00:02:c9:02:00:00:2b:45 brd
> 00
> :ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
> 
> But I see the same info when I turn back on the same IB port!
> 
> 4: ib0: <BROADCAST,MULTICAST,UP> mtu 2044 qdisc pfifo_fast
>     link/[32] 
> 00:00:04:04:fe:80:00:00:00:00:00:00:00:02:c9:02:00:00:2b:45 brd
> 00
> :ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
> 
> The only way I can see if the link is up is to look for 
> "RUNNING" in the
> ifconfig output.
> 
> 
> 
> 
> ------- You are receiving this mail because: -------
> You reported the bug, or are watching the reporter.
>