I can't believe I'm saying this, but maybe this is a job for NWAM profiles?
DTrace along with MDB can help you debug your NIC driver problems. I wish I could just say what you need to probe.. :-) You might also try different ACPI modes if it seems to be an interrupt problem. Mike On Sat, 2012-10-20 at 18:22 +0400, Jim Klimov wrote: > I really wonder how DTrace can help here - what should I track? ;) > > 2012-10-20 18:07, Michael Stapleton wrote: > > IPMP does not have to be configured with test addresses. IPMP will uses > > "Link based" by default if your NIC drivers support link state > > notification. > > The documentation really needs work... I think the documentation shows > > examples of "Probe based" because it is much more difficult to configure > > than "Link based". > > Simply add the interfaces to the same IPMP group. Done.. > > Yes, thanks. Today I stumbled upon this nice post about "simple > IPMP setup" without active probing, might help me. I wonder why > these are not wrapped in dladm management ;) > > http://cooperlees.com/blog/?p=328 > > And these are examples of "complicated setup" with active probing > which I was reluctant to do, from Joerg Moellenkamp's excellent > posts: > > http://www.c0t0d0s0.org/archives/6298-Less-known-Solaris-features-IP-Multipathing-Part-6-New-IPMP.html > http://www.c0t0d0s0.org/archives/6300-Less-known-Solaris-features-IP-Multipathing-Part-8-Classic-IPMP.html > > > > But, IPMP is really about uptime, not band width. If you want to use > > multiple interfaces, they all need IP addresses. They just do not have > > to have "Test" addresses. > > > > To use Aggregations, your switch and device drivers must support it. > > I think your NIC drivers have to be GLDv3 compliant. > > In fact, as per my original post, I don't exactly need IPMP nor > LACP for my case. My laptop has a single wired NIC which can be > represented by different drivers, one at a time. I want to make > switching between these simple and transparent to other network > configs on the system (vnics, zones and such). > > I plan to try "simple IPMP" and/or vanity naming (i.e. try to > name all of the possible NIC names "net0") and see if that works > later this weekend. > > At the moment I still wonder if any of these work while a device > is absent altogether (i.e. when the dual-booted OI is on physical > hardware, it has rge or gani but no e1000g; when it's booted as > a VM, it has e1000g but no rge/gani). I am now concerned that aggr > does not seem to work at all if a component device is missing - > this does seem like a bad bug in architecture or implementation. > > Thanks, > //Jim Klimov > > > > > _______________________________________________ > OpenIndiana-discuss mailing list > OpenIndiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss _______________________________________________ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
